Mohammad Ali Khojastepour

Outage minimization with limited feedback for the fading relay channel

In this paper, we consider practical methods to approach the theoretical performance limits in the fading relay channel under different assumptions of transmitter channel knowledge. Specifically, we consider two degrees of transmitter channel knowledge: 1) perfect feedback is available and power control is employed and 2) no channel state knowledge is available at the transmitters and only spatial power allocation is possible. First, when perfect feedback is available, the optimal power control policy determines the ultimate limits of performance for constant rate transmission in the slow fading environment. However, in practice, perfect channel knowledge is not possible at the transmitters due to the finite capacity of the feedback links. We find practical methods to approach this performance limit through the use of power control with finite rate feedback. The finite-rate feedback results are shown for the low-complexity, full-diversity amplify-and-forward (AF) protocol. Interestingly, we see that only a few feedback bits are needed to achieve most of the gains of the optimal perfect feedback power control algorithm. Second, we consider the performance limit when the transmitters have no channel state knowledge and derive the optimal spatial power allocation between the source and relay for a given sum power constraint for the AF protocol. For most practical cases of interest, equal power allocation between the source and relay is shown to be nearly optimal. Our work suggests that there is minimal power savings from using spatial power allocation at the transmitters. To obtain large performance improvements over constant power transmission, it is imperative to have feedback for each realization of the channel state to allow for temporal power control.

Outage minimization and optimal power control for the fading relay channel

In this work, we show that in the wireless relay network, a tremendous savings in energy can be achieved by having side information at the transmitters and by employing power control. We present efficient protocols and the corresponding optimal power control policies that approach the universal lower bound on the outage probability of the block fading relay channel. Each of the proposed protocols have their own utility for specific channel conditions. However, a hybrid protocol between two known coding schemes is the best scheme for all channel conditions and is sufficient to approach the lower bound on outage probability. Unlike the single link channel, we show that exploiting the knowledge of the channel at the transmitters can significantly lower the outage even if the transmit powers at the source and relay have to be kept constant. In this case, it is also demonstrated that the lower bound on outage is closely followed by the outage probability of the hybrid protocol. Our results reveal that exploiting the right network protocol in conjunction with power control result in orders of magnitude savings in power over direct transmission for a target performance level.

Bounds on achievable rates for general multi-terminal networks with practical constraints

We consider the problem of communication in a general multi-terminal network where each node of the network is a potential sender or receiver (or both) but it cannot do both functions together. The motivation for this assumption comes from the fact that current radios in sensor nodes operate in TDD mode when the transmitting and receiving frequencies are the same. We label such a radio as a cheap radio and the corresponding node of the network as a cheap node. We derive bounds on the achievable rates in a general multi-terminal network with finite number of states. The derived bounds coincide with the known cut-set bound [11] of network information theory if the network has just one state. Also, the bounds trivially hold in the network with cheap nodes because such a network operates in a finite number of states when the number of nodes is finite. As an example, application of these bounds in the multi-hop network and the relay channel with cheap nodes is presented. In both of these cases, the bounds are tight enough to provide converses for the coding theorems [16], and thus their respective capacities are derived.

The case for antenna cancellation for scalable full-duplex wireless communications

Recent works have considered the feasibility of full duplex (FD) wireless communications in practice. While the first FD system by Choi et.al. relied on a specific antenna cancellation technique to achieve a significant portion of self-interference cancellation, the various limitations of this technique prompted latter works to move away from antenna cancellation and rely on analog cancellation achieved through channel estimation. However, the latter systems in turn require the use of variable attenuator and delay elements that need to be automatically tuned to compensate for the self-interference channel. This not only adds complexity to the overall system but also makes the performance sensitive to wide-band channels. More importantly, none of the existing FD schemes can be readily scaled to MIMO systems. In this context, we revisit the role of antenna cancellation in FD communications and show that it has more potential in its applicability to FD than previously thought. We advocate a design that overcomes the limitations that have been pointed out in the literature. We then extend this to a two-stage design that allows both transmit and receive versions of antenna cancellation to be jointly leveraged. Finally, we illustrate an extension of our design to MIMO systems, where a combination of both MIMO and FD can be realized in tandem.

Delay-constrained scheduling: power efficiency, filter design, and bounds

In this paper, packet scheduling with maximum delay constraints is considered with the objective to minimize average transmit power over Gaussian channels. The main emphasis is on deriving robust schedulers which do not rely on the knowledge of the source arrival process. Towards that end, we first show that all schedulers (robust or otherwise) which guarantee a maximum queuing delay for each packet are equivalent to a time-varying linear filter. Using the connection between filtering and scheduling, we study the design of optimal power minimizing robust schedulers. Two cases, motivated by filtering connection, are studied in detail. First, a time-invariant robust scheduler is presented and its performance is completely characterized. Second, we present the optimal time-varying robust scheduler, and show that it has a very intuitive time water-filling structure. We also present upper and lower bounds on the performance of power-minimizing schedulers as a function of delay constraints. The new results form an important step towards understanding of the packet time-scale interactions between physical layer metric of power and network layer metric of delay

Cross-layer optimization for streaming scalable video over fading wireless networks

We present a cross-layer design of transmitting scalable video streams from a base station to multiple clients over a shared fading wireless network by jointly considering the application layer information and the wireless channel conditions. We first design a long-term resource allocation algorithm that determines the optimal wireless scheduling policy in order to maximize the weighted sum of average video quality of all streams. We prove that our algorithm achieves the global optimum even though the problem is not concave in the parameter space. We then devise two on-line scheduling algorithms that utilize the results obtained by the long-term resource allocation algorithm for user and packet scheduling as well as video frame dropping strategy. We compare our schemes with existing video scheduling and buffer management schemes in the literature and simulation results show our proposed schemes significantly outperform existing ones.

LDPC Code Design for Half-Duplex Cooperative Relay

The authors consider the design of LDPC codes for cooperative relay systems in the half-duplex mode. The capacity of halfduplex relay channels has been studied previously but the design of good channel codes for such channels remains a challenging problem. Employing an efficient relay protocol, we transform the half-duplex relay code design problem into a problem of ratecompatible LDPC code design where different code segments experience different SNRs. The density evolution with conventional Gaussian approximation for single user channels, which assumes invariant SNR within one codeword, is not capable of accurately predicting the code performance for this system. Here we develop a density evolution with a modified Gaussian approximation that takes into account the SNR variation in one received codeword as well as the rate-compatibility constraint. We then optimize the code ensemble using a modified differential evolution procedure. Extensive simulations are carried out to demonstrate that the proposed algorithm offers more accurate prediction of code performance in half-duplex relay channels than the conventional methods, and the optimized codes achieve a significant gain over existing codes.

LDPC-coded cooperative relay systems: performance analysis and code design

We treat the problem of designing low-density parity-check (LDPC) codes to approach the capacity of relay channels. We consider an efficient analysis framework that decouples the factor graph (FG) of a B-block transmission into successive partial FGs, each of which denotes a two-block transmission. We develop design methods to find the optimum code ensemble for the partial FG. In particular, we formulate the relay operations and the destination operations as equivalent virtual MISO and MIMO systems, and employ a binary symmetric channel (BSC) model for the relay node output. For AWGN channels, we further develop a Gaussian approximation for the detector output at the destination node. Jointly treating the relay and the destination, we analyze the performance of the LDPC-coded relay system using the extrinsic mutual information transfer(EXIT) chart technique. Furthermore, differential evolution is employed to search for the optimum code ensemble. Our results show that the optimized codes always outperform the regular LDPC codes with a significant gain; in the AWGN case, when Protocol-II is employed and the relay is close to the source, the optimized code performs within 0.1dB to the capacity bound.

Code design for the relay channel and factor graph decoding

Recent information theoretical results have shown a considerable improvement in the performance of communication systems through the use of relaying and cooperation. Despite a collection of strong information theoretical results on the relay channel, there has been almost no attention to real code design for the relay channel. In this paper, we present a powerful modular code design approach for the relay channel and corresponding decoding algorithms based on the factor graph representation of the code. For most of the relay channel conditions, the constructed codes with a low-complexity simple relay protocol outperforms any possible code design for the direct channel by achieving an E/sub b//N/sub 0/ below the minimum required E/sub b//N/sub 0/ of single-link transmission. Moreover, the designed codes achieve a gap of less than 1 dB (at a BER of 10/sup -6/) to the Shannon limit for the relay channel with a code length of only 2 /spl times/ 10/sup 4/ bits.

The capacity of average and peak power constrained fading channels with channel side information

We derive the ergodic capacity of discrete-time fading channel with additive Gaussian noise subject to both peak and average power constraint. The average power can be interpreted as the cost that we incur to achieve a certain rate. On the other hand, the motivation of this analysis comes from the fuel that there is also a peak power limitation in practical communication system. It is been shown that the optimal power adaption is no longer water-filling or constant power adaption which is the case where there is no limitation on the peak power. The numerical results show that the importance of peak power constraint becomes negligible for relatively low available average power, while it is limiting the capacity to be finite even as the available average power goes to infinity.