Adaptive Resource Allocation in OFDMA Relay-Aided Cooperative Cellular Networks

Abstract

Adaptive resource allocation to exploit multiuser diversity and spacial diversity in OFDMA Relay-aided cooperative cellular networks is studied. Assuming that the BS has all the channel state information (CSI), an optimization problem for subcarrier assignment, relay selection and power allocation that maximizes downlink capacity of the system is formulated. Since the optimal solution is complex to obtain, a suboptimal solution dividing the problem into two sub-problems is proposed. Assuming equal power allocation in BS and each relay, the first sub-problem for joint subcarrier assignment and relay selection is a combinatorial optimization problem, for which an iterative algorithm derived from necessary optimality condition is proposed. Given the results from the first subproblem, the second sub-problem for power allocation becomes a convex optimization problem and can be solved by an iterative method which optimizes the power of BS and each relay separately. Numerical results showed that two iterative algorithms can converge in reasonable steps and achieve significant improvement on system capacity due to two diversity schemes and power fully exploited.

Mixed Time-Scale Generalized Fair Scheduling for Amplify-and-Forward Relay Networks

“…Step 3 finds the sub-channel coupling and user that maximizes the gradient decent of the objective function, according to (9). After finding the best sub-channel coupling, the allocation for that user is increased and the user's rate is updated in Step 4 and Step 5.…”

Dynamic Resource Allocation with Threshold in OFDMA-based Relay Networks

“…To avoid this drawback, a solution of resource allocation, which divides the optimization problem into two sub-problems, is proposed [11]- [12] . In [11], an iterative algorithm is proposed to solve the first sub-problem: an optimization problem of combining subcarrier assignment with relay selection. The second sub-problem of power allocation can be solved by another iterative method.…”

“…The second sub-problem of power allocation can be solved by another iterative method. Similar to [11], in [12], optimization problem is also divided into two sub-problems: (1) subchannel allocation; (2) routing and power allocation, both of which are solved by an iterative approach respectively. Although the performance of these two algorithms is improved, computational complexity is still high owing to their iterative two-step approaches.…”

“…Although the performance of these two algorithms is improved, computational complexity is still high owing to their iterative two-step approaches. Beside, the studies of subcarrier allocation in [11] and [12] are based on an assumption that the subcarriers used by a relay node in the source-relay (SR) link are the same as those in the relay-destination (RD) link. In fact, the best subcarrier in the SR link may not be the best one in the RD link.…”

Genetic algorithm based adaptive resource allocation in OFDMA system for heterogeneous traffic

“…In fact, a lot of research efforts have been devoted to construct the optimization model, and more efforts are focused on reaching the optimal or near-optimal solution. Many convex optimization methods are well employed, for example, the well-known Lagrange multiplier [14] to obtain a local optimization with equal or unequal constraints, KTT conditions [15] to decide a localized maximization for a regular point, and dual algorithm is conducted to dissolve the combination optimization into two-layer problem [16]. However, most of the convex optimization problems desire tight or elastic requirement to the objective function and feasible regions which is not practicable for the nonconvex utility function since many QoS oriented functions are concave shape-like contrarily.…”