Optimal Relay-Subset Selection and Time-Allocation in Decode-and-Forward Cooperative Networks
📝 Abstract
We present the optimal relay-subset selection and transmission-time for a decode-and-forward, half-duplex cooperative network of arbitrary size. The resource allocation is obtained by maximizing over the rates obtained for each possible subset of active relays, and the unique time allocation for each set can be obtained by solving a linear system of equations. We also present a simple recursive algorithm for the optimization problem which reduces the computational load of finding the required matrix inverses, and reduces the number of required iterations. Our results, in terms of outage rate, confirm the benefit of adding potential relays to a small network and the diminishing marginal returns for a larger network. We also show that optimizing over the channel resources ensures that more relays are active over a larger SNR range, and that linear network constellations significantly outperform grid constellations. Through simulations, the optimization is shown to be robust to node numbering.
💡 Analysis
We present the optimal relay-subset selection and transmission-time for a decode-and-forward, half-duplex cooperative network of arbitrary size. The resource allocation is obtained by maximizing over the rates obtained for each possible subset of active relays, and the unique time allocation for each set can be obtained by solving a linear system of equations. We also present a simple recursive algorithm for the optimization problem which reduces the computational load of finding the required matrix inverses, and reduces the number of required iterations. Our results, in terms of outage rate, confirm the benefit of adding potential relays to a small network and the diminishing marginal returns for a larger network. We also show that optimizing over the channel resources ensures that more relays are active over a larger SNR range, and that linear network constellations significantly outperform grid constellations. Through simulations, the optimization is shown to be robust to node numbering.
📄 Content
Cooperation has become a popular technique to implement diversity in the absence of multiple antennas at receiving and transmitting nodes [1]- [3]. In this context, resource allocation in cooperative networks has recently become an active research area, and has been investigated under many scenarios and metrics.
In this paper, we address the problem of resource allocation, in terms of channel resources (time or bandwidth), in multi-relay networks with arbitrary connections. We describe the contributions of the paper in detail after a brief review of the pertinent literature.
For the single-relay case, several works have dealt with various aspects of resource allocation, in terms of power and/or bandwidth and time. Yao et al. determine the optimal power and time allocation for relayed transmissions specifically in the low-power regime [4]. Larsson and Cao present various strategies for allocating power and channel resources under energy constraints [5]. For the channel resource allocation problem, however, the authors consider selection combining only and do not address the scenario of joint decoding of the source and relay signals. The works in [6]- [8] address the problem of power and channel resource allocation under sum average power constraints. Optimal time and bandwidth allocation using instantaneous and average channel conditions is obtained using power control in [9]. Channel resource allocation under fixed power is developed in [10] In networks with multiple relays, the available literature can be classified into two groups: networks where relays do not communicate with one another (parallel-relay networks), and networks without restrictions on relay communication (arbitrarily-connected networks). Resource allocation for the former has been addressed in [11]- [14]. Ibrahimi and Liang develop the optimal power allocation for a multi-relay cooperative OFDMA amplify-and-forward (AF) system [11]. By maximizing the channel mutual information, Anghel et al. find the optimal power allocation for multiple parallel relays in AF networks [12], [13]. A more general solution is given in [14] where the authors give the optimal power and channel resource allocation for a parallel-relay network with individual power constraints on the nodes.
To the best of our knowledge, the problem of channel resource allocation for arbitrarily-connected networks and dedicated multiple access has not been addressed in the literature. In general, works in the area of multi-relay systems with arbitrary links generally neglect the bandwidth penalty arising from multiple hops by assuming either full-duplex nodes, a bandwidth-unconstrained system, or the availability of channel phase information at the transmitter [15]- [27]. These assumptions, however, are not realistic for practical wireless networks, where nodes are likely to be half-duplex, phase information is very difficult to obtain at the transmitter, and bandwidth is a scarce resource. To fill this void, in this paper we investigate the problem of resource allocation in a bandwidth-constrained, cooperative, decode-and-forward (DF), wireless network, and consider the most general setting where multiple relays can transmit can cooperate with each other in transmitting information between source and destination. In this setting, we address the joint problem of optimal selection of a relaying subset and allocation of time resources to the selected relays. The resource allocation is framed in the context of mesh networks of relatively simple and inexpensive nodes. We concentrate on resource allocation in terms of transmission time only, removing power allocation from the optimization; we further simply the problem by considering orthogonal transmissions. This is motivated by the need to reduce complexity, allowing for nodes which can implement the resource allocation simply by switching on and off. For a system without power allocation, a solution to this problem provides an upper bound on cooperative performance in multi-relay network where dedicated channels are assigned for each source transmission.
To the best of our knowledge, no other work provides a solution to time-allocation for an arbitrarily connected cooperative network. The solution can be interpreted as a generalization of the opportunistic protocol presented by Gunduz and Erkip, where the relay is active only when it increases the outage rate [6]. In terms of the resource allocation solution, it is also a generalization of the solution in [10], where channel resource allocation is determined under fixed power for a three-node DF network. The solution can also be interpreted as a generalization of node selection [3], [28]- [30] under relaxed transmission constraints, where transmission can occur on multiple time-slots and relays can communicate with one another.
This paper is structured as follows. Section II describes the system model. In Section III and IV we develop the proposed resource allocation scheme an
This content is AI-processed based on ArXiv data.