WO2016205765A1

WO2016205765A1 - Approaches for high speed global packet data services leo/meo satellite systems

Info

Publication number: WO2016205765A1
Application number: PCT/US2016/038260
Authority: WO
Inventors: Channasandra Ravishankar; John Corrigan; Rajeev Gopal; Yash Vasavada; James Jong; Nassir BENAMMAR; Gaguk Zakaria; Anthony Noerpel; Harish Ramchandran; Xiaoling Huang
Original assignee: Hughes Network Systems, Llc
Priority date: 2015-06-17
Filing date: 2016-06-17
Publication date: 2016-12-22

Abstract

A satellite system comprises LEO satellites and MEO satellites, and a control plane protocol architecture. The PHY, MAC, MAC/RLC and RRC layers are optimized for satellite environment. When the satellites are not processing satellites, eNB functions are implemented in a satellite gateway, and, when the satellites are processing satellites, protocol architecture in the control plane differ from LTE, as follows: PHY layer is moved to the communicating LEO/MEO satellite on the user link, MAC/RLC, RRC and PDCP are be located in satellite or gateway depending on satellite complexity, and the need to have mesh connectivity between UTs. When the RRC is implemented in the satellite, the RRC is divided into RRC-Lower and RRC-Upper layers. The RRC-L is satellite-based, and handles UT handover. The RRC-U is eNB-based, and handles resource management functions. The RRC-U communicates with the PDCP layer in the eNB to configure security, header and data compression.

Description

APPROACHES FOR HIGH SPEED GLOBAL PACKET DATA SERVICES FOR LEO/MEO SATELLITE

SYSTEMS

BACKGROUND

[1] Terrestrial communication systems continue to provide higher and higher speed multimedia ( e.g., voice, data, video, images, etc.) services to end-users. Such services (e.g., Third Generation (3G) and Fourth Generation Long Term Evolution (4G LTE) systems and services) can also accommodate differentiated quality of service (QoS) across various applications. To facilitate this, terrestrial architectures are moving towards an end-to-end all-Internet Protocol (IP) architecture that unifies all services, including voice, over the IP bearer. In parallel, mobile satellite systems are being designed to complement and/or coexist with terrestrial coverage depending on spectrum sharing rules and operator choice. With the advances in processing power of portable computers, mobile phones and other highly portable devices, the average user has grown accustomed to sophisticated applications (e.g., streaming video, radio broadcasts, video games, etc.), which place tremendous strain on network resources. Further, such users have grown to expect ubiquitous global coverage. The Web as well as other Internet services rely on protocols and networking architectures that offer great flexibility and robustness; however, such infrastructure may be inefficient in transporting Web traffic, which can result in large user response time, particularly if the traffic has to traverse an intermediary network with a relatively large latency (e.g., a satellite network). Such high mobility, enhanced processing power of devices, and growth of low-latency applications, however, puts an immense strain on current terrestrial and satellite communications systems.

[2] What is needed, therefore, is an approach for a low earth orbit (LEO) / medium earth orbit (MEO) multi-satellite communications system for efficiently providing high speed and high quality packet data services.

SOME EXAMPLE EMBODIMENTS

[3] Embodiments of the present invention advantageously address the foregoing requirements and needs, as well as others, by providing an approach for providing high speed and high quality packet data services via a LEO/MEO satellite system. The LEO/MEO satellites may be processing satellites. When LEO/MEO satellites are processing satellites, IP packets and Layer 2 frames transmitted by user terminals are recovered at the satellite and transmitted on the gateway links and/or inter-satellite links. Similarly, in the direction from network to user terminal, IP packets and Layer 2 frames transmitted by gateways are recovered at the satellite and transmitted on the user links. The frequency and format of transmission on gateway and user links may be different. In addition, the transmission to and from user terminal on a user link may be different. Similarly, the transmission to and from gateway on a gateway link may be different. The architecture also permits transmission from user terminal to another user terminal directly without traversing through a gateway. Similarly, the architecture permits direct gateway to gateway communication via the satellite constellation. When LEO/MEO satellites are not processing satellites (i.e., they are bent-pipe satellites), communication is directly between user terminal and gateway with a frequency translation between gateway links and user links.

[4] In accordance with example embodiments, an overall network architecture is shown in FIG. 1. The user terminal (UT) may be in one of a multiplicity of beams in the user link. Satellites, and therefore beams corresponding those satellites move (for satellite-fixed beams) over the user terminal as the LEO/MEO constellation moves even if the user terminal is not moving. Accordingly, beam-to-beam and satellite-to-satellite handover are required in this scenario. User terminals are typically equipped with a tracking antenna that is preferably electronically steered. However, the design does not preclude terminals using mechanical steering. In another embodiment, the satellite attempts to steer its antenna such that beams remain in the same place on the earth surface (also called earth-fixed beams). In this case, there is no need for beam-to-beam handover. The system also supports gateway to gateway handover to cater to cases where a user terminal may be in motion and it crosses from one gateway region to another. Gateway to Gateway handover would also be necessary when a Gateway fails or when the capacity of the gateway is such that it cannot accept any additional sessions. As part of the above mentioned beam-to-beam, satellite-to-satellite and gateway-to-gateway handovers, frequency handovers occur in a multiple frequency reuse system. To this end, the system design also supports frequency handover even when there is no beam-to-beam, satellite- to-satellite and gateway-to-gateway handovers; this will be the case when a frequency is deemed unusable due to interference and/or when it is required to move a terminal to a different frequency for resource usage efficiency issues and for services such as IP multicast.

[5] Certain system features are as follows:

• Powerful FEC coding, near theoretical channel performance;

• Adaptive Coding & Modulation (ACM) improves throughput every channel condition;

• Power-conserving design reduces power to enable battery/solar powered user terminal (sleep/wake paging cycle);

• MAC layer design for efficient Bandwidth-on-Demand;

• Support for Small and Large terminal types as well as fixed and mobile terminal types including Aeronautical and Maritime terminals;

• Quality-of-Service (QoS) support for multiple service types;

• Simplified satellite design to minimize technical and costs risks;

• Simplified routing/switching function in the satellite using a centralized route determination function in each gateway that determines optimal routes. This removes the burden for satellite to be dynamically figuring out the routes;

• Mobility Management functions enable beam, satellite, gateway and frequency handovers;

• Scalable Gateway architecture to cater to different throughputs and different number of LEO/MEO satellites that it would need to communicate with;

• Standard wireless and network protocols to utilize commercial implementations and evolution;

[6] Still other aspects, features, and advantages of the present invention are readily apparent from the following detailed description, simply by illustrating a number of particular embodiments and implementations, including the best mode contemplated for carrying out the present invention. The present invention is also capable of other and different embodiments, and its several details can be modified in various obvious respects, all without departing from the spirit and scope of the present invention. Accordingly, the drawing and description are to be regarded as illustrative in nature, and not as restrictive.

BRIEF DESCRIPTION OF THE DRAWINGS

[7] Embodiments of the present invention are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings, in which like reference numerals refer to similar elements, and in which:

[8] FIG. 1 illustrates the high-level architecture of a low earth orbit (LEO) / medium earth orbit (MEO) satellite system, according to example embodiments;

[9] FIG. 2A illustrates the user plane protocol architecture of a 4G long-term evolution

(LTE) terrestrial system;

[10] FIG. 2B illustrates the user plane protocol architecture for a LEO/MEO satellite system, according to example embodiments;

[11] FIG. 3A illustrates the control plane protocol architecture of a 4G long-term evolution (LTE) terrestrial system;

[12] FIG. 3B illustrates the control plane protocol architecture for a LEO/MEO satellite system, according to example embodiments;

[13] FIG. 4A illustrates a physical layer burst, according to example embodiments;

[14] FIG. 4B illustrates a return link RACH window, according to example embodiments;

[15] FIG. 4C illustrates an example of control messages (such as RACH and traffic) multiplexed on the same narrowband carrier, according to example embodiments;

[16] FIG. 4D illustrates an example of synchronous HARQ signaling, according to example embodiments;

[17] FIG. 4E illustrates an example of hybrid HARQ signaling, according to example embodiments; [18] FIG. 5A depicts a 3-D graph showing antenna gain in a satellite system, according to example embodiments;

[19] FIG. 5B depicts a graph showing system throughput with adaptive coding and modulation in a satellite system, according to example embodiments;

[20] FIG. 6A illustrates modification of a DVB-S2 format to achieve the establishment of frame markers reflecting system time for a user terminal battery conservation technique, according to example embodiments;

[21] FIG. 6B illustrates a user terminal state behavior for reduction in terminal battery usage, according to example embodiments;

[22] FIG. 6C illustrates satellite transmissions to a user terminal for reduction in terminal battery usage, according to example embodiments;

[23] FIG. 7A illustrates example of carrier channels for access and traffic channels of a flexible media access control (MAC) layer bandwidth on demand design, according to example embodiments;

[24] FIG. 7B illustrates an example MAC layer bandwidth on demand unsolicited uplink grant (UUG), according to example embodiments;

[25] FIG. 8 illustrates the RLC MAC layer context of a satellite system, according to example embodiments;

[26] FIG. 9 illustrates an example MAC layer downlink burst and public information (PUI), according to example embodiments;

[27] FIG. 10 illustrates an example MAC layer block boundary encoding, according to example embodiments;

[28] FIG. 11A illustrates an example slot assignment, according to example embodiments;

[29] FIG. 11B illustrates an example terminal addressing, according to example embodiments; [30] FIG. 12A illustrates the RRC control plane architecture of a satellite system, according to example embodiments;

[31] FIG. 12B illustrates an example RRC state diagram, according to example embodiments;

[32] FIG. 12C illustrates RRC user terminal identifiers for identifying and tracking UT context, according to example embodiments;

[33] FIG. 12D depicts a signal flow diagram illustrating RRC connection establishment, according to example embodiments;

[34] FIG. 12E depicts a signal flow diagram illustrating an RRC attach and bearer setup procedure, according to example embodiments;

[35] FIG. 12F depicts a signal flow diagram illustrating an RRC transition to idle mode, according to example embodiments;

[36] FIG. 12G depicts a signal flow diagram illustrating RRC EPC originated paging, according to example embodiments;

[37] FIG. 12H depicts a signal flow diagram illustrating an RRC user terminal originated MAC activation, according to example embodiments;

[38] FIG. 121 depicts a signal flow diagram illustrating an RRC SAN originated paging, according to example embodiments;

[39] FIG. 13A depicts a signal flow diagram illustrating a route management determination, according to example embodiments;

[40] FIG. 13B illustrates an example end-to-end routing protocol stack structure, according to example embodiments;

[41] FIG. 13C illustrates an example APN I P network domain structure, according to example embodiments; [42] FIG. 13D illustrates an example satellite interface domain structure, according to example embodiments;

[43] FIG. 13E illustrates an example gateway (GW) network domain structure, according to example embodiments;

[44] FIG. 14A depicts a signal flow diagram illustrating call flow messaging for initial registration and subsequent data transfer call phases, according to example embodiments;

[45] FIGs. 14B and 14C illustrate a routing change (not a handover) occurring in the satellite due to handover on the gateway link, according to example embodiments;

[46] FIG. 14D illustrates a user terminal mobility management context structure, according to example embodiments;

[47] FIG. 15A depicts a signal flow diagram illustrating a beam to beam handover, according to example embodiments;

[48] FIG. 15B depicts a signal flow diagram illustrating the preparation phase of a satellite to satellite handover, according to example embodiments;

[49] FIG. 15C depicts a signal flow diagram illustrating the data transfer phase of a satellite to satellite handover, according to example embodiments;

[50] FIG. 15D depicts a signal flow diagram illustrating the preparation phase of a gateway to gateway handover, according to example embodiments;

[51] FIG. 15E depicts a signal flow diagram illustrating the data transfer phase of a gateway to gateway handover, according to example embodiments;

[52] FIG. 15F illustrates a gateway to gateway handover, where the user terminal is at one gateway, according to example embodiments;

[53] FIG. 15G illustrates a gateway to gateway handover, where the user terminal moves to another gateway, according to example embodiments; [54] FIG. 15H depicts a signal flow diagram illustrating the preparation phase of a gateway to gateway handover with a satellite change, according to example embodiments;

[55] FIG. 151 depicts a signal flow diagram illustrating the data transfer phase of a gateway to gateway handover with a satellite change, according to example embodiments;

[56] FIG. 16A illustrates a diagram of the return link timing and frame numbering synchronization, according to example embodiments;

[57] FIG. 16B illustrates a switching time for user terminal transition from receive to transmit, according to example embodiments;

[58] FIG. 16C illustrates synchronization for half duplex operation where the satellite coverage area on the ground is divided into multiple Half Duplex Divisioned (HDD) groups, according to example embodiments;

[59] Fig. 16D illustrates gateway timing synchronization, according to example embodiments;

[60] Fig. 16E illustrates a gateway frame numbering scheme for synchronization, according to example embodiments;

[61] Fig. 17A illustrates an example gateway architecture, according to example embodiments; and

[62] Fig. 17B illustrates interfaces to an e-node B function of a gateway, according to example embodiments.

DETAILED DESCRIPTION

[63] System architectures and associated processes for providing high speed and high quality packet data services via a LEO/MEO satellite system are described. In the following description, for the purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the invention. It is apparent, however, that the invention may be practiced without these specific details or with an equivalent arrangement. In other instances, well-known structures and devices are shown in block diagram form in order to avoid unnecessarily obscuring the invention.

[64] As will be appreciated, a module or component (as referred to herein) may be composed of software component(s), which are stored in a memory or other computer-readable storage medium, and executed by one or more processors or CPUs of the respective devices. As will also be appreciated, however, a module may alternatively be composed of hardware component(s) or firmware component(s), or a combination of hardware, firmware and/or software components. Further, with respect to the various example embodiments described herein, while certain of the functions are described as being performed by certain components or modules (or combinations thereof), such descriptions are provided as examples and are thus not intended to be limiting. Accordingly, any such functions may be envisioned as being performed by other components or modules (or combinations thereof), without departing from the spirit and general scope of the present invention. Moreover, the methods, processes and approaches described herein may be processor-implemented using processing circuitry that may comprise one or more microprocessors, application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other devices operable to be configured or programmed to implement the systems and/or methods described herein. For implementation on such devices that are operable to execute software instructions, the flow diagrams and methods described herein may be implemented in processor instructions stored in a computer-readable medium, such as executable software stored in a computer memory store.

[65] Further, terminology referring to computer-readable media or computer media or the like as used herein refers to any medium that participates in providing instructions to the processor of a computer or processor module or component for execution. Such a medium may take many forms, including but not limited to non-transitory non-volatile media and volatile media. Non-volatile media include, for example, optical disk media, magnetic disk media or electrical disk media (e.g., solid state disk or SDD). Volatile media include dynamic memory, such random access memory or RAM. Common forms of computer-readable media include, for example, floppy or flexible disk, hard disk, magnetic tape, any other magnetic medium, CD ROM, CDRW, DVD, any other optical medium, random access memory (RAM), programmable read only memory (PROM), erasable PROM, flash EPROM, any other memory chip or cartridge, or any other medium from which a computer can read data.

[66] Various forms of computer-readable media may be involved in providing instructions to a processor for execution. For example, the instructions for carrying out at least part of the present invention may initially be borne on a magnetic disk of a remote computer. In such a scenario, the remote computer loads the instructions into main memory and sends the instructions over a telephone line using a modem. A modem of a local computer system receives the data on the telephone line and uses an infrared transmitter to convert the data to an infrared signal and transmit the infrared signal to a portable computing device, such as a personal digital assistance (PDA) and a laptop. An infrared detector on the portable computing device receives the information and instructions borne by the infrared signal and places the data on a bus. The bus conveys the data to main memory, from which a processor retrieves and executes the instructions. The instructions received by main memory may optionally be stored on storage device either before or after execution by processor.

ARCHITECTURE

[67] FIG. 1 illustrates the high-level architecture of a low earth orbit (LEO) / medium earth orbit (MEO) satellite system, according to example embodiments. As illustrated by Figure 1, the Ku band in user-link and V/Q band in Gateway link as examples. Other frequencies that are mutually exclusive may also be used in Gateway link and user links. As further shown in Figure 1, Satellite Gateways are connected via terrestrial links or via the existing LEO/MEO satellite constellation or via a GEO satellite system. IP Core network resembles that of a classical 4G-LTE network with the Border Gateway playing the role of PDN-Gateway (PGW) of LTE core network. Other elements that have a correspondence to 4G LTE core network include Subscription server (equivalent to the Home Subscription Server -HSS), Management Server (equivalent of MME) and Security Server (equivalent to AuC). Although the Serving Gateway (SGW) is not explicitly shown, it is expected to be part of the Satellite Gateway and/or PGW.

[68] FIG. 2A illustrates the user plane protocol architecture of a 4G long-term evolution (LTE) terrestrial system, and FIG. 2B illustrates the user plane protocol architecture for a LEO/MEO satellite system, according to example embodiments. It is noted from Figure 2B that the interface from eNB' to SGW is a standard Sl-U interface. This feature permits the use of COTS core network element. Similarly, all interfaces within the core network and to/from core network are based on 4G LTE standards.

[69] FIG. 3A illustrates the control plane protocol architecture of a 4G long-term evolution (LTE) terrestrial system, and FIG. 3B illustrates the control plane protocol architecture for a LEO/MEO satellite system, according to example embodiments. Similar to user-plane discussion above, this has resemblance to the control plane 4G LTE protocol architecture (Figure 3a) with the following additional key differences. Satellite system protocol architecture for Control Plane is similar to terrestrial 4G-LTE Control Plane protocol architecture shown in Figure 3A. The PHY, MAC/RLC and RRC layers are optimized for satellite environment. When satellites involved are not processing satellites, the eNB functions are implemented in a satellite gateway. However, for systems that have processing satellites, protocol architecture in Control Plane for the satellite system have the following key differences.

• PHY layer is moved to the communicating LEO/MEO satellite on the user link.

• MAC/RLC, RRC and PDCP may be located in satellite or gateway depending on permitted complexity of satellite, the need to have mesh connectivity between user terminals. The entity in the Gateway performing these functions is called herein as eNB'.

• When RRC is implemented in satellite, RRC is divided into RRC-L (RRC-Lower) and RRC-U (RRC-Upper) layers; RRC-L is located in the satellite and is responsible for handover signaling with UT. RRC-U is located in the eNB' and is responsible for resource management functions including admission control.

• RRC-U communicates with PDCP layer in eNB' to configure security, header compression and data compression schemes. ❖ In Figure 3B, although Ku band is depicted for the user link, system design embodiments also facilitate use of Ka band or L/S bands for improved availability.

[70] It is noted from Figure 3B that the interface from eNB' to SGW is a standard Sl-U interface. Similarly, from Figure 3B, the interface from eNB' to MME is a standard Sl-MME interface These features permit the use of COTS core network element. Similarly, all interfaces within the core network and to/from core network are based on 4G LTE standards.

PHYSICAL LAYER

[71] On the user link, use of multi-carrier (each with its own Power Amplifier) and single wideband carrier are considered for this system. Single wideband carrier provides better resource usage (multiplexing) efficiency, however will require a power amplifier with higher output power. Use of multiple carrier with its own power amplifier solves this problem, however, analog multiplexing to an antenna port as well as resource multiplexing inefficiency have to be managed. Physical layer is based on state-of-the-art LDPC codes or turbo codes in both forward and return links. The following are some attributes of forward link and return link multiple access schemes.

❖ Forward Link

- FDMA/TDM

- Control/traffic channels multiplexed in time

❖ Return Link

- Traffic channels: MF-TDMA or terrestrial standard such as SC-FDMA

- Control channels/narrow band data: Slotted Aloha, spread spectrum with MUD etc.

[72] Return Link Physical and Logical Channels are as follows:

❖ Return Link Physical and Logical Channels are as follows:

- Packet Data Channel (PDCH): Transport PDTCH (Packet Data Traffic Channel) logical channel and Uplink Control Messages - Random Access Channel (RACH): Transport RACH channel

The following table provides a description of the logical channels in return link

Forward Link Physical and Logical Channels are as follows:

- Packet Data Channel (PDCH): Transport PDTCH (Packet Data Traffic Channel) logical channel and Downlink Control Messages

- Packet Control Channel (PCCH): Transport following logical channels (Broadcast Channel, Paging Channel, Access Grant Channel, Packet Data Control Channel)

[75] The following table provides description of the logical channels in forward I

[76] Time slot and frame Definition Example:

- 1 ms frame consists of 12 time slots

- A time slot is a minimum duration that carries burst [77] FIG. 4A illustrates a physical layer burst, according to example embodiments.

• The Burst consists of o UW, Control Header (PUI), Payload o PUI is located at the front of the burst: Contains MCS, ULMAP, DLMAP, Payload presence information, ARQ related signaling, Power Control Command o UW symbols may be placed across burst

[78] For PUI, an alternate embodiment is to use convolutional codes instead of LDPC/Turbo Codes.

• A burst can carry either PDTCH only or PCCH/PDTCH o No difference in physical layer processing o The classification of PCCH/PDTCH is done in MAC o Burst contains PCCH/PDTCH is PI/2-BPSK modulated

• PCCH can be scheduled every 10 ms o UT may wake up every [10] ms to read PUI and check for PAGE (The presence of payload is indicated in PUI, minimizing Satellite and UT power consumption)

[79] A given burst may carry blocks that have different modulation and FEC schemes. If a burst were to contain blocks that were BPSK and QPSK modulated, for a processing satellite it is proposed to repeat the QPSK symbols instead of changing modulation scheme to BPSK within a burst but perform appropriate repetition for improved performance . This keeps satellite design simple. In other words, if one of the payloads were to be carried using pi/2 BPSK and other payloads were to be carried using pi/4 QPSK, then the BPSK symbols of the payload would be carried using repeated QPSK symbols so that the entire burst would look like pi/4 QPSK. The receiver then combined multiple pi/4 QPSK symbols for the BPSK region of the burst.

[80] Frame Hierarchy Exa - Time slot: 1/12 ms (burst duration in multiple integer of TS)

- Frame: 1 ms (single or multiple Uplink/downlink allocation signaling)

- Multi-frame: 10 ms (paging once every 10 ms)

- Super frame: 40 ms (BCCH info update with class 1 information occurring every 10 ms)

- Hyper frame: 1 second

[81] Forward Link Transmission structure example

- Extremely simple and efficient for signaling

- Efficient reference signal design that serves multiple purposes o Fast acquisition, reliable synchronization, etc.

- Dedicated control field with predefined number of bits available every 1/12 ms for any essential control signaling such as downlink, uplink allocation, MCS, DRX, hybrid ARQ feedback, etc.

- Shared PDCH available every 1/12 ms which could be periodically (every 10 ms) used for providing control info simultaneously to many users (i.e., broadcasting system information)

RACH Design in Return Link

[82] FIG. 4B illustrates a return link RACH window, according to example embodiments. The RACH is carried on a narrowband carrier with large RACH window to accommodate timing uncertainties within a beam. In the single burst format for narrow band PDCH and RACH only additional preamble is attached at the beginning and end of the burst when RACH is transmitted, which allows uniform design of PDCH and RACH design.

[83] FIG. 4C illustrates an example of control messages (such as RACH and traffic) multiplexed on the same narrowband carrier, according to example embodiments. [84] Further control messages (such as RACH and traffic) can be multiplexed on the same narrowband carrier. Traffic on narrowband carrier allows for small packets such as TCP ACK to be transmitted and at the same time off-loading the wider band traffic carriers.

Hybrid ARQ

[85] Given that the delays in LEO/MEO satellites are much shorter than that of GEO satellites, concepts such as Hybrid ARQ (HARQ) are applicable in this system. HARQ Parameters are derived based on the following assumptions:

❖ Return Link : 20 Msym/sec. Burst duration in return link is 0.5 ms

❖ Processing delay is 2 ms,Propagation delay between Terminal and Satellite 4 ms

❖ Synchronous transmission

❖ ACK sent immediately on receiving the packet ( Stop and Wait)

❖ If only one burst is sent, next opportunity to retransmit the burst is after 13 ms ( 26 return link time units).

Hence a total of 26 HARQ processes assumed using return link HARQ framework similar to LTE. Signaling for Synchronous HARQ

[86] Additional control signaling from SAT to UT indicate retransmission or new transmission. If synchronous transmission is employed, then retransmissions are scheduled exactly HARQ_window slots after the original transmission. If an allocation is made in uplink slot k, then slot k is for original transmission, slot k+ HARQ_window is for retransmission 1, slot k+ 2*HARQ_window is for retransmission 2. Only when ACK received can the same HARQ process (and corresponding time slots) can be assigned to other allocations.

[87] FIG. 4D illustrates an example of synchronous HARQ signaling, according to example embodiments. If too many allocations are made in the HARQ_window then system might be unable to accommodate new allocations. Therefore a threshold limit is set on the number of new transmissions within a HARQ_window. Asynchronous HARQ

❖ Restrictions imposed by synchronous HARQ from scheduling point of view

❖ Is it possible to use asynchronous HARQ

- Explicit process ID needed

• Process ID width will be the number of HARQ process supported o Process ID is 3 bits for LTE FDD

- How long to store packets in the buffer ?

❖ Asynchronous new transmissions and asynchronous retransmissions

- Buffer storage can be potentially huge

- Need limits on maximum number of HARQ processes supported Hybrid Hybrid ARQ (HHARQ)

[88] FIG. 4E illustrates an example of hybrid HARQ signaling, according to example embodiments. The hybrid Hybrid ARQ (HHARQ) facilitates use of asynchronous transmissions for new transmissions and scheduling retransmissions in a synchronous manner. The buffer size depends on the time interval between the original transmission and the retransmission. An explicit HARQ process ID is used to achieve this

Bundling

[89] When users are in disadvantaged location, instead of waiting for ACK/NACK from satellite, bundle the new transmission and the retransmissions. When retransmitted packet contents are similar to original packet (Chase combining) then this is similar to repetition. This may introduce time interval between original transmission and retransmissions to take advantage of time diversity.

Dynamic Link Adaptation

[90] The satellite links will support dynamic link adaptation to cater to different channel conditions that the user terminal experiences as the satellite moves. The extent to which beam pattern gains change as the satellite moves depends on the satellite antenna shape, aperture size and frequency of operation. Figure below indicates that the gain can decrease much as 8-10 dB at the edge of the beam as compared to the center of the beam for about a 20 cm antenna horn operating at 14.5 GHz frequency.

[91] FIG. 5A depicts a 3-D graph showing antenna gain in a satellite system, according to example embodiments.

[92] In addition, environmental conditions (rain, clouds etc.) also influence the attenuation that the signal undergoes. Finally, the C/l that the links have are a function of the load and in the return link the relative location and power of interfering users compared to the user of interest. All these factors motivate the need for dynamic link adaptation whereby the throughput is maximized for a given channel condition by changing the modulation and coding scheme based on channel condition. This is illustrated in figure below.

[93] FIG. 5B depicts a graph showing system throughput with adaptive coding and modulation in a satellite system, according to example embodiments.

[94] To this end, the physical layer bearers are defined such that they span the various modulation and coding schemes to give 10 to 15 dB of channel quality variation. The modulation schemes are from repeated pi/2 BPSK to pi/4 QPSK to 3pi/4 8-PSK to 16-APSK. Coding rates vary from Rate 1/5 code to Rate 9/10 code using LDPC codes.

Battery Conservation Feature

[95] Battery conservation is an important feature for both user terminal as well as satellite. When a user terminal is battery or solar powered, this feature becomes even more critically important. Accordingly, a scheme is developed whereby a user terminal in idle mode will only wake up for less than 0.5 ms every 10 ms, thereby leading to a reduction in battery consumption by up to 95%. This is facilitated by the satellite/network only paging the user terminal at the wake up time of the terminal. Therefore, there is a need to establish the concept of a system time. The system time is established by the satellite based on the GPS receiver it has. It uses the 1 pulse per second ticks to establish frame markers in the satellite downlink. The user terminal is made aware of these ticks by using a special sequence of unique words in the user link downlink at the 1 PPS or an integer submultiple of 1 second. An example scheme is illustrated in figure below wherein a DVB-S2 format is modified to achieve this establishment of frame markers. This concept is also applicable to framework other than DVB-S2 such as the one described earlier in the document for Physical Layer.

[96] FIG. 6A illustrates modification of a DVB-S2 format to achieve the establishment of frame markers reflecting system time for a user terminal battery conservation technique, according to example embodiments. User terminals performs dual hypothesis to determine the frame marker used by the satellite. An important advantage of this scheme is that, since these markers are based on GPS, these markers are aligned across all satellites - this helps when there is a satellite to satellite handover, since there is no need to re-establish these frame markers. User terminal states and behavior is shown below that achieves about 95% reduction in battery usage.

[97] FIG. 6B illustrates a user terminal state behavior for reduction in terminal battery usage, according to example embodiments.

[98] The air interface described herein further reduces battery consumption in user terminal even in the Connected State. Here the downlink burst in the front has information regarding the user(s) to which the rest of the burst belongs. A user terminal first demodulates and decodes the front part of the burst to determine if the rest of the burst had data for it or not. If not, user terminal powers down until the beginning of next burst thereby saving power. As an example, when the front portion of the burst is 4 microseconds long and burst duration is about 80 microseconds, then assuming that the terminal is consuming power for no more than 8 microseconds, the saving is 90% in connected mode when there is no burst for it. If a terminal is active 20% of the time and 10% of the downlink bursts are meant for the terminal, then total battery consumption in connected mode is only is 7.8% leading to a power saving of 92.2% for the baseband part of the user terminal.

[99] A facility for further reduction in battery power consumption in idle mode is also provided. Here the front part of the burst also contains a bit to indicate whether the rest of the burst contains a page message or not. User terminal therefore needs to wake up for about 8 microseconds every 10 ms in idle mode leading to a very negligible power consumption in idle mode.

[100] In a LEO/MEO environment, the satellite orbits the earth in less than 2hours and will therefore have very short visibility to the sun for charging its battery in each orbit. Therefore, if there are no active users at all in a beam, the satellite does not spend any power for 10 ms. Every 10 ms satellite would transmit system information for about 80 microseconds leading to a battery saving of more than 99%. In other words, power overhead will be less than 1%.

[101] FIG. 6C illustrates satellite transmissions to a user terminal for reduction in terminal battery usage, according to example embodiments. Based on extrapolation of timing and frequency, user terminal will be able to go to complete sleep for about 10 ms and wake up and still be able to demodulate and decode the burst.

MAC Layer Attributes

[102] A flexible and efficient MAC layer is provided for both access as well as traffic. MAC layer design is based on Bandwidth on Demand (BoD) and therefore the proposed method provides dynamic allocation of resources based on demand. For access channels, a narrowband (e.g., 1 MHz carriers) channels are used, and for traffic, narrowband or wideband (e.g., 17.5 MHz) channels are used in a dynamic way so as to use the available uplink resources in a very efficient manner.

[103] FIG. 7A illustrates example of carrier channels for access and traffic channels of a flexible media access control (MAC) layer bandwidth on demand design, according to example embodiments. Key attributes include:

• Low bandwidth (1 MHz) access channel for initial access o Minimizes signaling overhead o Keeps satellite receiver simple o Can be used to request access as well as transfer data

• Frequent Uplink allocation signaling to minimize delay (0.5 ms) • Control and Data packet multiplexing for improved spectral efficiency

• Unsolicited uplink grants to further decrease delay of uplink allocation (see figure

below). Here the satellite provides an Unsolicited Uplink Grant (UUG) to user terminal so that the terminal can transmit TCP Acknowledgements using UUG on the traffic channel right away without soliciting resources from the satellite. This provides delay efficiency. It is noted that for half duplex terminals, the terminal cannot receive while it is transmitting. Therefore, satellite cannot transmit in downlink when user terminal has been allocated an uplink resource, including UUG. The UUG scheme is thus different for half-duplex terminals compared to full-duplex terminals. For half-duplex terminals, the frequency of UUGs will be less than that for full duplex terminals.

[104] FIG. 7B illustrates an example MAC layer bandwidth on demand unsolicited uplink grant (UUG), according to example embodiments.

[105] MAC layer signaling and scheduler design supports multiple terminal types. The half- duplex and full duplex terminal types mentioned above is one such example. It also supports fixed and mobile terminals, wideband and ultra-wideband terminals. Further, beam level signaling is employed instead of carrier level signaling. In beam level signaling, a user terminal is not tied to a particular carrier in uplink. Network may allocate resources on any uplink carrier dynamically based on availability and honoring the switching constraints of the user terminal to switch frequencies.

RLC/MAC Context

[106] FIG. 8 illustrates the RLC MAC layer context of a satellite system, according to example embodiments.

Downlink Burst and Public Information (PUI)

[107] FIG. 9 illustrates an example MAC layer downlink burst and public information (PUI), according to example embodiments.

[108] The PUI carries the following set of information:

• Modulation and coding scheme of the FEC blocks in the Private Information (PRI) • User identifier of the data carried in the FEC blocks

• Uplink allocation o User identifier o Power control bits o Carrier index o Number of assigned consecutive transmissions o Indication if additional information present in PRI

• Block Boundary

[109] This feature provides flexibility and fewer RLC/MAC headers, i.e., less overhead. It indicates how the MAC layer should interpret the multiple FEC blocks in the PRI. It indicates whether the MAC layer byte stream straddle the FEC blocks. The field indicates how the multiple FEC blocks are grouped as shown in the figure below.

[110] FIG. 10 illustrates an example MAC layer block boundary encoding, according to example embodiments.

Downlink User Addressing and Terminal Power Consumption

[111] The addressing scheme tries to minimize terminal battery usage and the processing of downlink data not destined to UT. It improves channel efficiency by filling up downlink bursts and sending to multiple terminals at the same time. It also provides low latency when scheduling terminals

Slot Assignment and Terminal Addressing

[112] FIG. 11A illustrates an example slot assignment, according to example embodiments.

• User terminal is provided o Slot Id and periodicity, i.e., slot#5 every 2 frames o Or a bitmap on 000000 000011. i.e., last 2 slots of every frame • Decision factors o Terminal type and battery status o User demand, specifically any guaranteed bit traffic flows o Number of active flows and aggregate user demand o More efficient to transmit to terminals with similar channel condition in the same burst

• Terminal bitmap can be updated and sent to UT based on above factors Terminal Addressing

[113] Embodiments provide for a mixture of the 2 operating points balancing throughput efficiency, low latency and terminal power efficiency

• Unicast Addressing: o A user identifier per FEC block (4 * 16 bit user identifier)

^■ a terminal only processes the FEC block when the address matches

• Multi-user Addressing o When assigning a slot to a terminal, a terminal is also provided an identifier x from (0-63), only valid in the assigned slot. Terminal address is 2^Λχ

[114] FIG. 11B illustrates an example terminal addressing, according to example embodiments. o Use of a 64bit user identifier bitmap to instruct UT to decode FEC blocks o Only terminals with bit set need to decode FEC blocks, i.e., if terminal address is 2^Λχ and bit x is set

[115] Terminal addressing space is 64 * Periodicity (slots) Radio Resource Management Protocol

[116] RRC is based on LTE procedures with modifications for the satellite environment. Use standard interfaces to the Evolved Packet Core(EPC) network (36.413 - S1AP, 36.423 - X2AP, 24.301 - NAS)

[117] FIG. 12A illustrates the RRC control plane architecture of a satellite system, according to example embodiments.

• RRC context o Resides in GW (SAN-C subsystem) and UT o Implements connection, bearer, mobility, resource management functions o Uses acknowledged mode bearers (PDCP, RLC-AM, MAC, PHY) o Also uses Common Control Channels for some procedures o Ciphering, integrity protection provided by PDCP o Carries NAS MM/SM signaling transparently o Interfaces at SI interface to LTE core (MME) RRC State Model

[118] Addition of RRC_PCH state enables UT dynamic DRX feature and reduces time to data transfer state

[119] FIG. 12B illustrates an example RRC state diagram, according to example embodiments.

[120] The RRC states and sub states can be categorized as follows • RRC Idle

• UT control plane not established

^■ No UT context in SAN ^■ UT context exists in EPC (MME) if EMM-Registered

• UT monitors BCCH, PCH

• Connection establishment can be triggered by UT or by paging fro RRC Connected

• Control and data bearers context established

^■ UT context exists in EPC and in SAN a) RRC_ACTIVE

^■ Radio resources assigned

^■ UT monitors assigned slots in DL traffic channel

^■ UL/DL transmission can happen b) RRC_PCH

^■ No radio resources assigned

^■ UT monitors BCCH, PCH

^■ MAC layer reactivation can be triggered by UT or by paging from SAN

UT Identifiers

[121] Multiple UT identifiers are needed to identify and track UT context at different levels and in different states

[122] FIG. 12C illustrates RRC user terminal identifiers for identifying and tracking UT context, according to example embod

• SAN-assigned o RRC:

S-RNTI (Satellite Radio Network Temporary Id)

- GWID (8b) + Logical SAN-C Id [7b] + UT Id [17b] - TBC o MAC: ^■ MAC User id (16 bits)

^■ FL slot# (1-12), period (1-4?) and Slot-specific UT id (6 bits) SI Broadcast and CCCH Usage

• BCCH: System Information (SI)

o At any time, SI in a beam is broadcast by a single GW

^■ This primary GW assignment changes as the satellite moves over different GW's

• RACH

o RACH capacity can be shared by multiple GWs

o RACHs arriving in a beam may be destined to different GWs

^■ Default (primary) GW decides correct destination based on UT type and location

^■ UT remembers assigned GW

^■ See "GW Redirection" message flow for details

• AGCH, PCH

o AGCH and PCH capacity can be shared by multiple GWs

• Capacity is divided between GWs based on configuration

RRC Functions and Proced

Function Procedure jMessag Direction Bearer jSystem Information

jSystem Information broadcast

Ibroadcast !BCCH segments DL !SRBO

!DL Information Transfer DL Information Transfer DL SRB2

NAS information IUL Information Transfer UL Information Transfer UL SRB2 transfer Piggybacking NAS messages on certain RRC

messages

jPaging SRB0

RRC Connection Establishment

Request UL SRBO jl mmed iate Assig nment DL SRBO

RRC Connection Establishment

Immediate Assignment Reject DL SRBO jR RC C o nn e cL i o n S e i. u p D L SRBl

RRC Connection Setup Complete UL SRBl

RRC Connection

jRRC C onn ecti on Relea se jRRC Connection Release DL SRBl Management, MAC Activation Request UL SRBO Bearer Management MAC Activation

jMAC Activation Confirm DL SRBl

Security Mode Command DL SRBl

RRC Security Mode Security Mode Complete UL SRBl

Security Mode Failure UL SRBl jR RC C o nn e cLi qn Rec n f i g u r a ti o n D L SRBl jRRC Connection Reconfiguration RRC Connection Reconfiguration

Complete UL mm® jPosition yerification Req u est UL SRBO jPosition Verification

jPosition and Battery Position Verification Notify DL SRBO j Re porting jR RC Positi on Report UL SRBl jRRC Position Report

RRC Position Report Confirm DL iiiiii

RRC Connection Establishment

[123] UT uses this procedure to attach to the network on power-on

[124] FIG. 12D depicts a signal flow diagram illustrating RRC connection establishment, according to example embodiments.

• UT performs satellite and beam selection and camps on the selected beam.

• UT transmits RRC Connection Request on RACH. If the UT already knows its GW label (SAN- RC), it adds it, else it adds an empty destination label for the satellite to fill. The UT also supplies its GPS coordinates and S-TMSI (if available).

• Satellite receives the RACH and examines the GW ID in the label. If null, it fills in the default GW label to be used for RACH (points to the SAN-RC). The satellite also adds the source info identifying the (satellite, beam, carrier, frame, slot) and forwards to the SAN-RC (via ISL and the RFT). • SAN-R processes the RACH. This includes checking the UT coordinates in the RACH to determine/confirm the home GW and determine the TA (using GWSA), fetching the UT satellite/beam trajectory (using EDF), determining the SAN-C instance responsible for this UT (load balancing). SAN-RC then forwards the request with all this data to the SAN-C.

• SAN-C performs connection admission and creates the UT context, stores the position and trajectory, assigns the UT identifier (S-RNTI). SAN-C sends the Immediate Assignment on AGCH to the UT via the satellite. The Immediate Assignment contains the timing correction, forward timeslot assignment and MAC id, GW labels for RACH and traffic (SAN-R).

• SAN-C sends the RRC Connection Setup message on SRB1 via RFT and satellite to the UT.

This contains the MAC layer parameters for the UT, and other parameters needed in connected state. This includes information about upcoming beam & satellite handovers.

• UT sends the RRC Connection Setup Complete message on SRB1 with selected PLMN Id, GUMMEI, and a piggybacked NAS message. SAN-C sends the NAS message to the MME.

• SAN-C sends an SI INITIAL-UE message containing the UT's local SI AP Id, NAS message, the selected PLMN Id, GUMMEI, TAI, etc.

• The NAS procedure proceeds. See next chart. Attach and Bearer Setup

[125] UT completes attach and default bearer is established

[126] FIG. 12E depicts a signal flow diagram illustrating an RRC attach and bearer setup procedure, according to example embodiments.

• SAN-C sends an SI INITIAL-UE message containing the UT's local SI AP Id, NAS ATTACH message, the selected PLMN Id, GUMMEI, TAI, etc. (see previous chart).

• MME performs the NAS Security Mode procedure (mutual authentication and key

exchange) with the UT. This is carried transparently in NAS Downlink /Uplink Transport messages over SI and over SRB1 in RRC DL/UL Information Transfer messages. • MME gives SAN-C the UT context info in the Sllnitial Context Setup Request message. This includes UT security keys and the details of default E-RAB to be set up. It also includes the NAS ATTACH ACCEPT message for the UT

• SAN-C computes the ciphering and integrity protection keys and configures the user plane (SAN-U PDCP). It also performs AS security activation using the RRC Security Mode procedure over SRB1.

• SAN-C performs bearer admission, establishes the radio bearers for the default E-RAB

indicated by the MME and for SRB2. The RRC Connection Reconfiguration procedure is used to set up RBs. The RRC Connection Reconfiguration includes the return SAN-RU label to be used for data traffic.

• SAN-C replies to the MME with the SI Initial Context Setup Response, indicating which

bearers were set up successfully or failed. Local GTP transport addresses & TEIDs for the successful bearers is also included.

• UT replies to the MME with NAS ATTACH COMPLETE. This is carried transparently over SRB2 in RRC UL Information Transfer and over SI in SI Uplink NAS Transport messages

Transition to Idle on Inactivity

[127] When the UT is idle for a few seconds, it transitions to RRC_PCH. When the UT is idle for many minutes, it transitions from RRC_PCH to RRC IDLE

[128] FIG. 12F depicts a signal flow diagram illustrating an RRC transition to idle mode, according to example embodiments.

• SAN-R detects TBF inactivity and signals to the UT that the TBFs are being released through MAC signaling.

• When the TBF Release has been acknowledged by the UT, the SAN-R removes the MAC layer state for the UT and informs the SAN-U. The SAN-U removes its references to the SAN- R for this UT and informs the SAN-C that the MAC has become inactive. The RBs, RLC and PDCP contexts remain. The SAN-C changes the RRC state to RRC_PCH. The SAN-RU reference is cleared, but the rest of the UT context is retained.

• On extended inactivity, the SAN indicates to the EPC (MME) that it wishes to release the UE context by sending a UE Context Release Request with cause "user inactivity". The MME sends a SI UE Context Release Command.

• The SAN triggers SAN originated paging.

• When the UT responds, the SAN sends the UT an RRC Connection Release. The UT

transitions to RRC Idle and sets its DRX interval to the default value.

• The SAN sends the MME a SI UE Context Release Complete. The SAN also stores the UT's S- TMSI and last reported position in a database for use in idle mode paging.

Paging Scenarios

[129] Paging principles o Paging applies in RRC Idle and RRC_PCH states o UT reports its position to SAN if it moves more than [50 km] o SAN determines satellites and beams to be used to page at UT's last known position

UT NAS States UT RRC State Paging Mechanism Comment ί EMM Deregistered ! RRC Idle 1 None No data bearers, hence no ECM Idle paging needed ί EMM Deregistered ; RRC Connected 1 None Transient state while attaching ECM Idle ί EMM Registered ! RRC Idle 1 EPC triggers paging by SAN stores UT's last reported ! ECM Idle i MME-assigned id (S- coordinates when in RRC Idle

TMSI). state to facilitate accurate

I SAN pages based on UT's paging. last reported

coordinates.

EMM Registered RRC Connected None Transient state while

ECM Idle connecting

EMM Registered RRC Connected: SAN triggers paging by SAN stores UT's last reported ECM Connected RRC_PCH SAN-assigned id (S-RNTI). coordinates in UT context

SAN pages based on UT's when in RRC_PCH state. last reported

coordinates.

EMM Registered RRC Connected: None Active radio resources, hence ECM Connected RRC_ACTIVE no paging needed

EPC Originated Paging

[130] When the EPC has data to send, it pages the UT to request RRC connection establishment.

[131] FIG. 12G depicts a signal flow diagram illustrating RRC EPC originated paging, according to example embodiments.

• While in ECMJDLE, the EPC (SGW) receives DL data for the UT and triggers the MME to initiate paging. The MME sends SI PAGING to all the SAN-C's responsible for the TA. The message includes the UT temporary ID (S-TMSI) and TA list (TAL).

• Only one of the paged SAN-C's will be selected to do the paging. The SAN-C looks up the last known location of that S-TMSI in the database and determines the satellites/beams which cover that location (by querying EDF).

• The SAN-C sends paging requests to the SAN-R for each target satellite/beam with the S- TMSI and default DRX interval for the UT. • SAN-R schedules paging messages on the PCH slots for the target UT depending on the current DRX interval and sends them to the UT via the RFT and satellite.

• The UT is monitoring its paging opportunities based on the default DRX interval. It receives the S-TMSI based page and responds with a RRC Connection Establishment Request with cause "Paging Response". The RRC Connection Establishment completes and the UT initial context is set up.

• The NAS Service Request is conveyed to the MME within the SI Initial UE message at the conclusion of the RRC Connection Establishment.

• The NAS Service Request procedure takes place. This is similar to the Attach procedure in that the NAS authentication and security procedures take place and the bearers are established.

• When the E-RAB for the required EPC bearer context is established, the EPC (SGW) sends the queued downlink data that triggered paging.

UT Originated MC Activation

[132] When the UT has data to send, it requests MAC activation and transitions to RRC_ACTIVE

[133] FIG. 12H depicts a signal flow diagram illustrating an RRC user terminal originated MAC activation, according to example embodiments.

• While in RRC_PCH, the UT needs to transmit uplink traffic so triggers the MAC Activation Request procedure. It transmits a MAC Activation Request on RACH with cause "UL data". The UT already knows its GW label (SAN-RC), so it adds it to the message.

• Satellite receives the RACH, adds the source info (satellite, beam, carrier, frame, slot) and forwards to the SAN-RC (via ISL and the RFT).

• SAN-RC processes the RACH, which includes looking up the TAI and beam/satellite

trajectory for the current UT position. The S-RNTI in the message contains the logical SAN-C id. SAN-RC forwards the request with all this data to this SAN-C. SAN-C sends the Immediate Assignment on AGCH to the UT via the satellite. The Immediate Assignment contains the timing correction, forward timeslot assignment and MAC id, GW labels for RACH and traffic (SAN-R).

• SAN-C Sends the MAC Activation Confirm message on SRB2 via SAN-U, SAN-R, RFT and

satellite to the UT. This contains the MAC layer parameters for the UT, and other parameters needed in connected state. This includes information about upcoming beam & satellite handovers.

• UT starts monitoring the downlink channel for its assigned slot. When it receives an uplink allocation, it sends the uplink data.

SAN Originated Paging

[134] When the SAN has data to send, it pages the UT to request MAC activation

[135] FIG. 121 depicts a signal flow diagram illustrating an RRC SAN originated paging, according to example embodiments.

• While in RRC_PCH, the GW receives DL data from the EPC (SGW). The SAN-U queues the data and requests the SAN-C to trigger paging.

• SAN-C determines the satellites/beams to be paged based on the last known UT position (by querying EDF). It sends paging requests to the SAN-R with the S-RNTI and current DRX interval for the UT.

• SAN-R schedules paging messages on the PCH slots for the target UT depending on the

current DRX interval and sends them to the UT via the RFT and satellite.

• The UT is monitoring its paging opportunities based on its current DRX interval. It receives the S-RNTI based page and responds with a MAC Activation Request with cause "Paging Response". The UT already knows its GW label (SAN-R), so it adds it to the message.

• Satellite receives the RACH, adds the source info (satellite, beam, carrier, frame, slot) and forwards to the SAN-R (via ISL and the RFT). • SAN-R processes the RACH, which includes looking up the TAI and beam/satellite trajectory for the current UT position. The S-RNTI in the message contains the logical SAN-C id. SAN-R forwards the request with all this data to this SAN-C. SAN-C sends the Immediate

Assignment on AGCH to SAN-RC. SAN-RC forwards it to the UT via the satellite. The

Immediate Assignment contains the timing correction, forward timeslot assignment and MAC id, GW labels for RACH (SAN-RC) and traffic (SAN-RU).

• UT starts monitoring the downlink channel for its assigned slot. The SAN-U sends the

queued downlink data that triggered paging.

Differentiated Quality Of Service

[136] System and Air Interface signaling supports multiple levels of QoS. Traffic classes supported consistent with that defined in 4G LTE standards o Conversational Class o Streaming Class o Interactive Class o Background Class

[137] Weighted Fair Queuing (WFQ) based scheduling algorithms is proposed for QoS differentiation. Consistent treatment needed for good Quality of Experience (QoE) across the network. This includes o UT-SAT link QoS o SAT-SAT QoS o SAT-GW QoS o Backbone QoS

Route Management

[138] Route management in the system is proposed to be based on a Route Determination Function (RDF) in the gateway. The RDF determines the route that a packet has to take through the constellation so that it is transmitted in the appropriate user link downlink that the user terminal is communicating on. The basic principles of route management is as follows and shown in Figure RM-1:

• UT provides its position to satellite

• Satellite forwards to appropriate Gateway based on UT position and/or Traffic Engineering rules

• UT position also provided to Gateway

• Gateway has binding between UT ID, its IP address (Sl-AP ID) and its last reported position

• In forward link, Gateway RDF determines the satellite that covers UT position

[139] FIG. 13A depicts a signal flow diagram illustrating a route management determination, according to example embodiments.

[140] Route management is inherently tied to mobility management. As user terminals handover from beam to beam and satellite to satellite, the RDF has to update its routes. Mobility Management and Handover are described next.

Route Management Protocol

[141] Routing Protocols: o All the routing/switching/forwarding mechanisms used to get packets between UT and PDN o Includes layers 2 & 3 [142] FIG. 13B illustrates an example end-to-end routing protocol stack structure, according to example embodiments.

APN IP Domain

o Refers to addressing used between UTs and servers in APN(s) o Private, public or combination

^■ Multiple APN domains isolated from each other and from GW internal IP domain

^■ UT can be connected to multiple APNs at once

- And simultaneously have an IP address in each one

^■ NAT can be deployed in EPC P-GW

^■ Interface to Internet/data center via BGP router o IPv4 or IPv6 o UT IP address assigned by P-GW o P-GW can assign from a private pool and perform NAT o Can assign pools per APN o Completely isolated from GW internal IP domain o User IP packets are tunneled over end-to-end EPC bearers o Endpoints are P-GW and UT

[143] FIG. 13C illustrates an example APN IP network domain structure, according to example embodiments.

GW IP Domain

• Refers to addressing used for IP communication between GW components

• Private IP address space, e.g. 10.x. x.x o Subdivided by site (GW, NOC, etc.) o Further subdivided by functional group (management, user data, signaling, etc.)

• Isolated from end-user (APN) address domain(s) o Firewalled access possible for management, support, etc. GW Ethernet Domain

• Refers to Ethernet switching layer used to interconnect GW components

• Subdivided into individual collision domains: o SIF (Satellite Interface) domain

^■ Used to interface to satellite constellation via RFTs

^■ Interconnects neighbor GWs by means of L2VPNs o Traffic VLANs

^■ VLANs used to segregate other types of traffic, e.g.

- SAN-R to SAN-U traffic

- SAN-U to EPC traffic

- EPC to APN traffic

- Management traffic

[144] FIG. 13D illustrates an example satellite interface domain structure, and FIG. 13E illustrates an example gateway (GW) network domain structure, according to example embodiments.

Mobility Management

[145] For mobility management, it is assumed that

• Every given orbit has to be seen by two or more geographically diverse gateways • For normal operation, while one gateway suffices, two gateways are required for situations where one of them fails or in a deep rain fade zone ] At power ON,

• if no prior constellation ephemeris information is stored in UT,

- UT scans for best possible signal quality within the [+/- 57 degree elevation]

- UT picks best satellite in view

- User terminal sends Channel Request on access channel with GPS location to the selected satellite after reading system information; it also sends measurement report of the satellites in view

- SAT determines destination gateway based on policy (location, regulatory, traffic engineering etc.)

- If destination gateway not in the orbit, o SAT sends it to closest Gateway o Gateway (call it radio gateway) determines if there exists a satellite in the constellation that the UT can reach the intended gateway (that also meets the signal quality criterion) o If so,

Radio Gateway sends back a message to User Terminal about re- attempting on a different satellite (provides necessary parameters) that helps reach the intended gateway (this implies that a given gateway has information regarding the geographical coverage area of neighboring gateway)

UT performs this request through the new satellite and registers via this gateway (call it registered gateway) o If not, Radio Gateway creates a tunnel to the intended Gateway and forwards the RRC Connection Request to the intended Gateway (this becomes the Registered Gateway)

RRC layer in Registered Gateway responds to UT via the Radio Gateway

Gateway ID provided to the UT is the registered gateway. UT is also informed of the "Via" Gateway in this case

Subsequent transmissions from UT will contain Via and Registered Gateway.

If SAT can reach the Registered Gateway directly, then SAT forwards it to the registered gateway

[147] FIG. 14A depicts a signal flow diagram illustrating call flow messaging for initial registration and subsequent data transfer call phases, according to example embodiments.

[148] FIGs. 14B and 14C illustrate a routing change (not a handover) occurring in the satellite due to handover on the gateway link, according to example embodiments. In this case no handover occurs in the user link. The respective figures illustrate the following points. o A given satellite beam may be providing service to two different gateways; o Need for satellite to take appropriate switching decisions o For normal traffic, this decision is aided by user terminal indicating the registered gateway in data that it transmits to satellite o For terminals that are registering or re-registering, satellite determines the intended gateway o Terminal remains connected to the gateway with which it has registered o In general, a satellite may communicate with two gateways simultaneously o There may be cases where a gateway is too low of an elevation angle; in that case cross link gets used to reach the intended gateway o While a UT is in session, the satellite may move such that the gateway it is communicating is no longer visible to the satellite; o I n this case, there is no beam handover, no satellite handover, but there is an impact to the switching logic o To accomplish this switch, satellites advertise reachability of Gateways to other satellites o I n addition, Gateways upload switching tables to satellite UT States

• UT states determine what services a UT can get:

• UT states are defined in

o RRC Layer

o Non-Access Stratum (NAS) Layer

• RRC-Layer State

^■ RRC-ldleRRC-Active

^■ RRC-PCH

• NAS Layer State

o Evolved Packet System (EPS) Mobility Management (EMM) state

o EPS Connection Management (ECM) state

• EMM State

o EMM-Deregistered

o EMM-Registered

• ECM State

o ECM-ldle

o ECM-Connected

UT Mobility Management Context

• UT coverage change due to constellation movement o Supported by Beam/Satellite HO • Movement of UT within and between service areas o Supported by Position Reporting, Tracking Area Update and GW-GW HO

• GW coverage change due to constellation movement o Not addressed by RRC procedures, but by satellite network layer (SNL) routing

[149] FIG. 14D illustrates a user terminal mobility management context structure, according to example embodiments.

Handovers

• Each Gateway knows the satellites trajectories

• Each Terminal reports its location

• Based on the terminal location, Gateway estimates at what time the HO should be executed o This can be done in advanced since Gateways knows the terminal location and the

satellite movements

o Gateways sends Sat/beam HO trajectory to UT so that UT knows when the HO will happen

• MME involves in Gateway-Gateway HO since this HO requires path switch between EPC entities and the new Gateway

o Gateway-gateway HO mostly happens for moving terminal such as air plane

There are several HO scenarios described here

• Beam to Beam handover

• Satellite to Satellite Handover

• Gateway to Gateway Handover

[150] Beam to Beam Handover: FIG. 15A depicts a signal flow diagram illustrating a beam to beam handover, according to example embodiments.

[151] Satellite to Satellite Handover: FIG. 15B depicts a signal flow diagram illustrating the preparation phase of a satellite to satellite handover, and FIG. 15C depicts a signal flow diagram illustrating the data transfer phase of a satellite to satellite handover, according to example embodiments. [152] Gateway to Gateway Handover (In this case, the satellite is the same, however gateway handover was necessary): FIG. 15D depicts a signal flow diagram illustrating the preparation phase of a gateway to gateway handover, and FIG. 15E depicts a signal flow diagram illustrating the data transfer phase of a gateway to gateway handover, according to example embodiments.

[153] High-level Gateway to Gateway Handover (with satellite change): FIG. 15F illustrates a gateway to gateway handover, where the user terminal is at one gateway, and FIG. 15G illustrates a gateway to gateway handover, where the user terminal moves to another gateway, according to example embodiments.

[154] Gateway to Gateway Handover (where the satellites also change) -The following uses X2 procedures similar to 4G LTE. Gateway to Gateway handovers involving SI interface procedures may also be invoked if X2 interface is not available: FIG. 15H depicts a signal flow diagram illustrating the preparation phase of a gateway to gateway handover with a satellite change, and FIG. 151 depicts a signal flow diagram illustrating the data transfer phase of a gateway to gateway handover with a satellite change, according to example embodiments;

Traffic Shaping in Gateway to Manage Buffers in Satellite

• UT position is known in Gateway

• For a fixed UT, the antenna gain pattern changes as the satellite moves

o there could be 6 to 10 dB of gain variation

• Depending on the gain at a user location, Adaptive Coding and Modulation gets invoked in the forward link to that user in Ku-band user link

• This implies that the forward link throughout for a given user varies as the satellite moves

• When the throughput is low, the satellite would therefore need to buffer

• To minimize buffering requirements at the satellite, Gateway "shapes" the traffic to a given user based on user location

• Gateway reduces the rate at which transmits data to a given UT if Gateway determines that the user-link satellite antenna gain is low and vice-versa

o This is possible since gateway knows UT location and Gateway also knows the forward link beam gain of the user-link satellite • This will therefore minimize buffering requirement at the user-link satellite Flow Control Between Satellite and Gateway

• Traffic shaping helps minimize buffering requirement based on UT location and user-link satellite forward link gain.

• However there will be cases where the user link throughput has to be throttled depending on non-deterministic factors such as rain

• In this case, the buffers would start to grow in the satellite

• To better manage the depth of queues in the satellite, the user-link satellite implements a simple flow-control mechanism with the Gateway

• Here the user-link satellite will transmit the soft equivalent of RNR to Gateway o This can be done on satellite basis, beam-by-beam basis or user-by-user basis

• When soft-RNR is received by the Gateway, the Gateway throttles the rate at which it

injects data towards user-link satellite

Synchronization

[155] The speed of the LEO/MEO satellite as observed from a location on the earth is high. This high speed of the satellite results in a large satellite motion induced Doppler. The goal of the proposed synchronization scheme is to compensate for the large Doppler offset by exploiting the deterministic nature of the Doppler component.

Synopsis of the proposed scheme:

• Synchronization task at the GW is aided by the knowledge available at the GW of the

LEO/MEO satellite ephemeris, and the positions of the GW and of the UT.

• Similarly, the UT is equipped with the knowledge of the LEO/MEO satellite ephemeris and its position as well.

• To achieve the latter, the LEO/MEO satellite ephemeris is broadcast by the GW on the

forward link [156] Compared to an alternate scheme in which the predictable and deterministic nature of the LEO/MEO satellite Doppler component is not taken advantage of, the proposed synch approach:

• incurs a small increase in the broadcast message overhead (due to the provision of

broadcasting the LEO/MEO satellite ephemeris),

• while providing an increased efficiency in o the UT's forward link signal acquisition and handover measurement processes, o the Paging messaging transmission from the GW and the reception at the UT, and o GW's return link signal acquisition and handover measurement processes [157] These increased efficiencies translate to

• Faster times at the UT and at the GW for signal acquisition and for satellite-to-satellite

handover,

• a quicker response by the UT to the GW's Paging messages,

• an improvement in the UT's battery life (e.g., since the UT in the idle mode can be in the sleep mode in between the assigned paging frames, and also because the satellite/cell search is more efficient)

• a reduction in the satellite power and bandwidth consumed to send Paging messages to the UTs, and

• a reduced complexity of the GW and the UT acquisition and tracking receivers [158] User link may be in one of the following states:

• Cold Start

o Limited availability of satellite ephemeris, resulting in large uncertainties in Doppler and timing

o May require at the transmitter either (i) dedicated pilot/FCCH channel, or (ii) a large preamble followed by a small message packet (e.g., RACH). Similarly a receiver design with large acquisition window is needed • Warm Start

o Satellite ephemeris is available but may not have been recently updated

o Links in partially synchronized state

• Steady State (Idle and Connected Mode Handovers)

o Accurate ephemeris, and estimates of delay and Doppler, is available on both the ends of the link

o Guard bands and acquisition windows are smallest [159] User Link - Forward link synchronization

• Satellite and the UT both have a GPS disciplined oscillator

o Frequency reference is locked to GPS

o Frame markers derived based on GPS 1 pps timing ticks

o UT continually estimates the downlink delay and Doppler using the LEO/MEO satellite ephemeris data

• Downlink Timing Acquisition at the UT

o By adding the estimated downlink delay to its GPS based 1 pps ticks, the UT derives an estimate of the downlink frame markers

o UT opens an acquisition window for the downlink frame timing around this estimated frame marker

^■ Acquisition window is largest at the cold start (may be continuous), smaller in the warm start, and smallest in steady-state.

• Downlink Frequency Acquisition at the UT

o By adding the estimated downlink Doppler to its GPS disciplined frequency reference, the UT derives an estimate of the downlink frequency

o UT opens an acquisition window centered at this estimated downlink frequency

^■ Acquisition window is largest at the cold start (may be continuous), smaller in the warm start, and smallest in steady-state

• After the initial acquisition, the downlink timing and frequency are continually tracked by the UT receiver [160] User Link - Return link synchronization

• UT adds Transmit Receive Offset (TRO) to the downlink timing and frequency to derive the uplink timing and frequency

• TRO is dependent on the user link RTD (Round Trip Delay) and RTDop (Round Trip Doppler) o TRO is initially self-computed by the UT given the knowledge of the satellite ephemeris and its GPS position

o A Closed Loop Correction (CLC) feature, such as the one in GMR-1, is necessary to

prevent run-off conditions due to inaccuracies in the self-computed TRO at UT

o If the UT receives a TRO correction from the satellite in the CLC message, it extrapolates this last received TRO using the ephemeris data

o Therefore, in the steady state, the UT's return link transmissions are synchronized at the satellite

• A medium acquisition window preamble and a large acquisition window preamble would be useful to serve "corner cases" (warm/cold starts)

o To serve these cases, satellite transmits the beam center delay and Doppler Frame Numbering and Synchronization

[161] FIG. 16A illustrates a diagram of the return link timing and frame numbering synchronization, according to example embodiments. The UT continually adjusts the Transmit Receive Offset (TRO) as shown.

Half Duplex Operation

[162] Half duplex terminals cannot simultaneously transmit and receive

• As shown below, a switching time Δυχ_→τχ is required for the UT to transition from the receive to transmit (and vice versa, Δ_τχ→κχ is needed to switch from transmit to receive)

• A scheme for synchronization and resource allocation with the above constraint for the half duplex terminals is proposed [163] FIG. 16B illustrates a switching time for user terminal transition from receive to transmit, according to example embodiments. The satellite coverage area on the ground is divided into multiple Half Duplex Divisioned (HDD) groups:

• Size of the HDD group is determined by the duration of the packets assigned to the UTs on downlink

• Return uplink resource is assigned to the terminals in a given HDD group such that

• Terminals at the edge of the HDD group nearest to the satellite end their uplink

transmission Δ_τχ→κχ seconds before the beginning of the downlink packet (see UT B Tx Timeline in the figure below)

• Terminals at the edge of the HDD group farthest to the satellite begin their uplink

transmission _{RX→ x} seconds after the end of the downlink packet (see UT A Tx Timeline in the figure below)

[164] FIG. 16C illustrates synchronization for half duplex operation where the satellite coverage area on the ground is divided into multiple Half Duplex Divisioned (HDD) groups, according to example embodiments. It can be shown that, for any given duration of the packet(s) assigned to the UT on the downlink, the proposed scheme maximizes the Half Duplex Resource Allocation Efficiency (HDRAE):

• HDRAE (0 < HDRAE < 1) is defined as the ratio of the time that the terminal is active (in either transmitting or receiving) over the interval of time that covers the UT's Rx - Tx - Rx transitions

• In the example on the previous slide, over the three units of time (0, 1 and 2) that cover UT's Rx - Tx - Rx transitions, the terminal is active over 2 units. Therefore, the HDRAE is 0.666. This is the maximum attainable HDRAE given that the downlink resource in this example occupies one unit of time. [165] Further, the HDRAE can be improved (such that it approaches the maximum value of unity) by increasing the duration of the packet resource assigned to the UT either on the downlink or on the uplink. Two alternative example implementations may be as follows:

• In one approach, the HDD groups are mapped to the formed beams on the ground

• Although this may lead to suboptimal HDD groups (especially for the beams larger in geographic size compared to the size of the HDD group), an advantage is that a common resource allocation rule can be applied for all the terminals in a given beam

• In an alternate approach, the HDD groups are defined independent of the beam. This requires the use of a "terminal-location-aware" scheduler at the Gateway

• The benefit of this approach is the use of optimally-defined HDD groups. A drawback is that this requires a more complex scheduler design at the GW (e.g., scheduler has to perform handover of a given terminal from one HDD group to the next as the satellite traverses in its orbit)

Gateway Synchronization Scheme

• On the forward link, the timing, frequency, and frame numbering of the frames transmitted by the GW are aligned at the satellite to a GPS-derived system reference

• GW continuously adjusts transmit timing and frequency of all the forward uplink

transmissions to each LEO/MEO satellite to compensate for the feederlink delay and Doppler.

• GW calculates the required transmission offsets from the ephemeris data, and applies the calculated offsets to a system-level GPS derived synchronization reference signal

[166] Fig. 16D illustrates gateway timing synchronization, and Fig. 16E illustrates a gateway frame numbering scheme for synchronization, according to example embodiments.

UT-UT Direct Sessions

[167] There may be some sessions that require direct communication between user terminals without passing through the gateway. In such a case, the Reassembly function of the RLC layer is selectively implemented in satellite for those sessions that require direct terminal to terminal communication without involvement of the Gateway. In this case, security contexts and keys have to be exchanged directly between two user terminals.

Gateway Architecture

[168] Gateway architecture is based on 4G LTE radio network and core network architecture, modified for satellite environment. Here a Gateway has visibility to a number of LEO/MEO satellites depending on the location of gateway. Each Gateway has a number of tracking antennas in the V/Q band. Fig. 17A illustrates an example gateway architecture, according to example embodiments.

[169] Tracking antennas have the necessary radio modulation and demodulation functions so that the baseband from multiple tracking antennas may be transported to eNB's via optical fiber link. This architecture therefore permits gateway diversity to mitigate rain propagation effects. A diverse set of tracking antennas may be placed several tens of miles away and the digital baseband signal can be hauled via fiber to the common eNB. SGW, PGW and MME are standard 4G LTE core network elements. As discussed earlier, a key component of the Gateway is the Route Determination Function (RDF) that is responsible for generating the appropriate labels for IP packets to be transmitted to user terminals communicating via the LEO/MEO constellation. This provides the centralized architecture providing clear separation between control and user plane functions. Various interfaces to the eNB' function of a given Gateway is shown in Figure below.

[170] Fig. 17B illustrates interfaces to an e-node B function of a gateway, according to example embodiments. The gateway tracking antennas may be steerable antennas or phased array antennas. For the case of phased array antennas, it is possible to have a single large array of antenna elements that form multiple beams tracking multiple satellites.

[171] While example embodiments of the present invention may provide for various implementations (e.g., including hardware, firmware and/or software components), and, unless stated otherwise, all functions are performed by a CPU or a processor executing computer executable program code stored in a non-transitory memory or computer-readable storage medium, the various components can be implemented in different configurations of hardware, firmware, software, and/or a combination thereof. Except as otherwise disclosed herein, the various components shown in outline or in block form in the figures are individually well known and their internal construction and operation are not critical either to the making or using of this invention or to a description of the best mode thereof.

[172] In the preceding specification, various embodiments have been described with reference to the accompanying drawings. It will, however, be evident that various modifications may be made thereto, and additional embodiments may be implemented, without departing from the broader scope of the invention as set forth in the claims that follow. The specification and drawings are accordingly to be regarded in an illustrative rather than restrictive sense.

Claims

CLAIMS WHAT IS CLAIMED IS:

1. A satellite communications system comprising:

one or more low earth orbit (LEO) satellites;

one or more medium earth orbit (MEO) satellites;

a control plane protocol architecture where physical (PHY), media access control (MAC), MAC radio link control (MAC/RLC) and radio resource control (RRC) layers are optimized for a satellite environment, wherein, when the satellites are not processing satellites, e- node-B (eNB) functions are implemented in a satellite gateway, and, when the satellites are processing satellites, protocol architecture in the control plane for the satellite system have the following differences from LTE:

• PHY layer is moved to the communicating LEO/MEO satellite on the user link,

• MAC/RLC, RRC and packet data control plane (PDCP) may be located in satellite or gateway depending on complexity of the satellite, and the need to have mesh connectivity between user terminals,

• when the RRC is implemented in satellite, the RRC is divided into RRC-Lower (RRC-L) and RRC-Upper (RRC-U) layers, wherein the RRC-L is located in the satellite and is responsible for handover signaling with user terminals, and the RRC-U is located in the eNB' and is responsible for resource management functions including admission control, and

• the RRC-U communicates with the PDCP layer in the eNB' to configure security, header compression and data compression schemes.