Feedback Strategies for Wireless Communication

Over the last years, the interest in high data rate for wireless transmission has significantly increased. During the past decade, significant contributions have been made in modulation coding and resource allocation and have led to the current systems. Future wireless communication systems should provide a wide range of services at a reasonable cost and sufficient quality of services, comparable to wireline technologies. An important direction to increase the rate and performance is to design precoding and post-coding schemes by exploiting the knowledge of the wireless channel conditions at the transmitter and receiver.

In this chapter, we present a brief overview of basic wireless communication systems. The chapter starts with a description of wireless communication channel models including MIMO channel models and dual-polarized antennas channel models. We then discuss about digital modulation and orthogonal frequency division multiplex. Finally, we provide a short introduction on diversity and spatial multiplexing gain in wireless communication systems and MIMO space time block codes.

In this chapter, we study the different feedback strategies for the case of a single user wireless communication system where both the transmitter and the receiver are equipped with a single antenna. After a review of the capacity of finite state and Rayleigh fading channel, we have studied the adaptive transmission over time and frequency where rate and power are adapted in order to maximize the spectral efficiency. We have shown that depending on the availability of the channel state information (CSI) at the transmitter (CSIT) and at the receiver (CSIR), the capacity can significantly increase by exploiting the time variation of the channel. We extend the study to consider the important class of frequency selective channel where the power can be optimally shared among the frequencies. We study the adaptive transmission over time and frequency where rate and power are adapted in order to maximize the spectral efficiency. The adaptive modulation and coding where a joint optimization of the coding rate and modulation is a practical scheme to approach capacity. We then study channel prediction at the transmitter to compensate the delay due to feedback link. While average spectral efficiency is not affected by the time delay, there is a significant degradation on the average bit error depending on the length of the prediction filter. For wideband channel, due to correlation of the channel in frequency, the amount of feedback can be reduced by performing data compression. Finally, we consider the Automatic Repeat Request (ARQ) schemes that allow to build a reliable data transmission using ACK/NACK feedback message and data retransmission.

In this chapter, we consider different feedback strategies for the case of a single user MIMO wireless communication system. We examine the case where only one stream is sent from the transmitter to the receiver corresponding to the MISO case or MIMO with receiver combining. Since the information is contained in the channel direction, we mainly consider the case where the normalized MISO channel is quantized, feedback and beamforming is applied at the transmitter. We see that the codebook composed of vectors can also be seen as a finite set of subspaces in the Grassmanian manifold and we derive the codebook criterion, bounds on performance and methods of construction of codebooks. For the MIMO systems where multiple streams are transmitted to the receiver, linear precoding extends the beamforming precoding by premultiplying the different transmitted data streams using a precoding matrix. In that case, the channel direction information is a unitary matrix and depending on the selected criterion the codebook design should be adapted. We study the different techniques to exploit the spatial, time and frequency correlation of the channel. For spatial correlation, we show that the codebook design should be adapted by performing a local packing. For time correlation, the same idea can be exploited. Similarly to correlated source coding, the class of differential feedback strategies are the most promising ones. Finally, in the case of selective MIMO channels, we show that it is possible to reduce the quantity of feedback

In this chapter, we focus on the different feedback strategies for the case of a multiuser wireless communication system where both the transmitter and the receivers are equipped with a single antenna. Firstly, an information-theoretic overview is provided assuming that the users channel state information (CSI) is fully known at the transmitter. Then, user scheduling algorithms are introduced by taking into account different criteria including opportunistic and proportional fairness. Finally, we study the reduced and limited feedback algorithms in detail and provide the performance results for both single carrier and multicarrier based multiuser systems.

In this chapter, we focus on the different feedback strategies for the case of a multiuser wireless communication system with multiple transmitter and receiver antennas. Firstly, we analyze the capacity of the multiuser MIMO system with single receive antenna for uplink and downlink by assuming that the full channel state information (CSI) is available at the transmitter. Secondly, we examine the precoding and user selection algorithms for MIMO multiuser systems with one and multiple receive antennas in flat fading and frequency selective wireless channels. Since optimal precoding using dirty paper coding has a prohibitively high computational complexity due to the associated encoding process, it is a great practical interest to design MIMO multiuser systems with low complexity and a minimum CSI requirement at the transmitter side. One suboptimal approach is to apply linear precoding schemes, such as zero forcing beamforming (ZF-BF) or minimum mean square error criterion. Multiuser MIMO wireless communication with ZF-BF requires a brute-force exhaustive search over all possible user sets and the complexity of an exhaustive search is prohibitive when the number of users is large. In order to decrease the complexity of this search, several suboptimal user scheduling algorithms have been designed. Generally, these algorithms fall into two categories: Capacity-based and Frobenius norm-based algorithm. Lastly, we show the effect of reduced and limited feedback information including user selection at the receiver side and quantization for both single carrier and multicarrier transmissions. We perform user selection at the user side since the users having a poor channel (low norm or/and interference) should not take part in the user selection algorithm, nor feedback their channel information. By using a self-discrimination criterion at the receiver side, it is possible to reduce the feedback load and the complexity of the user selection algorithm at base station. We show different user selection criteria and quantization strategies to reduce feedback load for both single and multicarrier communication systems.

The increasing demand for wireless multimedia has led to coordinated multicell transmission which can increase data rate and reduces outage in cellular systems by mitigating intercell interference (ICI). In order to mitigate ICI, adaptive power allocation and multiple antenna techniques are employed in the multicell networks. In this chapter, we describe limited and reduced feedback strategies for SISO and MISO based cooperative multicell framework in flat fading and frequency selective channels. Since the amount of feedback bits increases with the number of users and base stations, it is important to perform a selection at the user side. In addition to that, the cooperative networks are quite sensitive to the quality of the channel state information of serving and interfering base stations to eliminate inter-cell interference. We examine the limited feedback strategies for multiantenna multicell systems. The purpose is to quantize the CSI belonging to serving and interfering BSs by applying bit partitioning strategies based on different criteria. We illustrate the performance results for cooperative multicell systems with reduced and limited feedback strategies.

In this chapter, we consider feedback strategies for Long Term Evolution (LTE) and LTE-Advanced based wireless communications systems. We first give an overview on LTE standard including frame structure, transmission modes and reference signals for the different releases. We introduce respectively link adaptation considering and channel measurement reporting modes including channel quality information (CQI), precoding matrix index (PMI) and rank indicator (RI). Single user and multiuser MIMO strategies and multicell MIMO transmission schemes including static and dynamic coordination techniques are then reviewed. Finally, design criteria and codebook constructions for downlink and uplink transmission are described in detail.

We conclude this book by summarizing and highlighting its contribution in the field of feedback strategies for wireless communication. We will also discuss some open research topics that were identified during the writing of this book.