Laboratory of video transmission over 3G/4G networks

Participants

Ann Ukhanova, DTU Fotonik, PhD student/Researche - team-leader

Vitaly Grinko, SUAI, student

Introduction

Video communications, as already used in the internet, become the second largest mass service after voice telephony. However video telephony is assumed to be about 10% in 3G networks with voice telephony traffic 5 years after implementation.

With the most modern Video compression techniques, a fast changing video image can be compressed by approx. 10:1. Where the majority of the image is stationary as in video telephony less than 2% of the uncompressed bits need to be sent across the network, relating to a 50:1 to 60:1 compression. This encoding of changes in a video data stream is the essence of MPG ( ISO Motion Picture Experts Group ). The characteristic nature of this data stream would be most suitable to a Variable Bit Rate (VBR) bearer service. This means that the transmission over an unreliable channel should be corrected with the instruments accessible on the PHY and MAC layer. Below you can find some useful information about it.

Interfaces to the physical layer

The physical layer (layer 1) is the lowest layer in the OSI Reference Model and it supports all functions required for the transmission of bit streams on the physical medium. The physical layer interfaces the Medium Access Control (MAC) Layer and the Radio Resource Control (RRC) Layer as depicted in figure 1.

Figure 1. Interfaces with Physical Layer

Interface to MAC

The physical layer interfaces the MAC entity of layer 2. Communication between the Physical Layer and MAC is in an abstract way performed by means of PHY-primitives defined which do not constrain implementations.

The PHY-primitives exchanged between the physical layer and the data link layer provide the following functions:

• transfer of transport blocks over the radio interface;
• indicate the status of the layer 1 to layer 2.

Interface to RRC

The physical layer interfaces the RRC entity of layer 3 in the network.

Communication is performed in an abstract way by means of CPHY-primitives. They do not constrain implementations.

The CPHY-primitives exchanged between the physical layer and the Network layer provide the following function:

• control of the configuration of the physical layer.

The currently identified exchange of information across that interface has only a local significance to the UE orNetwork

Services and functions of the physical layer

General

The physical layer offers data transport services to higher layers. The access to these services is through the use of transport channels via the MAC sub-layer. The characteristics of a transport channel are defined by its transport format (or format set), specifying the physical layer processing to be applied to the transport channel in question, such as convolutional channel coding and interleaving, and any service-specific rate matching as needed.

The physical layer operates exactly according to the L1 radio frame timing. A transport block is defined as the data accepted by the physical layer to be jointly encoded. The transmission block timing is then tied exactly to this L1 frame timing, e.g. every transmission block is generated precisely every 10ms, or a multiple of 10 ms. A UE can set up multiple transport channels simultaneously, each having own transport characteristics (e.g. offering different error correction capability). Each transport channel can be used for information stream transfer of one radio bearer or for layer 2 and higher layer signalling messages. The multiplexing of these transport channels onto the same or different physical channels is carried out by L1.

Overview of L1 functions

The physical layer performs the following main functions:

• FEC encoding/decoding of transport channels;
• measurements and indication to higher layers (e.g. FER, SIR, interference power, transmission power, etc…);
• error detection on transport channels;
• multiplexing of transport channels and demultiplexing of coded composite transport channels;
• rate matching;
• mapping of coded composite transport channels on physical channels;
• modulation and spreading/demodulation and despreading of physical channels;
• frequency and time (chip, bit, slot, frame) synchronisation;
• closed-loop power control;
• power weighting and combining of physical channels;

L1 interactions with L2 retransmission functionality

Provided that the RLC PDUs are mapped one-to-one onto the Transport Blocks, Error indication may be provided by L1 to L2. For that purpose, the L1 CRC can be used for individual error indication of each RLC PDU. The L1 CRC will then serve multiple purposes:

• error indication for uplink macro diversity selection combining (L1);
• error indication for each erroneous Transport Block in transparent and unacknowledged mode RLC;
• quality indication;
• error indication for each erroneous Transport Block in acknowledged mode RLC.

Regardless of the result of the CRC check, all Transport Blocks are delivered to L2 along with the associated error indications.

This project also addresses the important issues of error control for video transmission over 3G. Based on the time-varying wireless channel conditions and the essential defects of the traditional hybrid ARQ for real-time service, the architecture of the channel-adaptive hybrid ARQ/FEC is discussed and an algorithm for encoder that automatically adjust the parity data length and the maximum number of retransmissions is investigated.

Figure 2. Framework of a multiuser cross-layer video transmission system over wireless networks.

Also with the advancement of video-compression technology and the wide deploymentof wireless networks, there is an increasing demand for wireless video communicationservices, and many design challenges remain to be overcome. We would like to discuss how to dynamically allocate resources according to the changing environmentsand requirements, so as to improve the overall system performance andensure individual quality of service (QoS). Specifically, we will consider two aspectswith regard to design issues: cross-layer design, which jointly optimizes resourceutilization from the physical layer to the application layer, and multiuser diversity,which explores source and channel heterogeneity for different users. We will studyhow to efficiently transmit multiple video streams, encoded by current and futurevideo codecs, over resource-limited wireless networks such as 3G/4G cellular systemand future wireless local/metropolitan area networks (WLANs/WMANs).