LAUSR

In the cognitive radio ad hoc networks (CRAHNs), the spectrum availability may change from time to time and hop by hop. Thereby, the performance analysis and optimization for CRAHNs become so intricate that pure divide-and-conquer strategies by layered principles strand. Focusing on the five-layer involved performance analysis of CRAHNs, we set up a cross-layer design framework mathematically and solve it through vertical decomposition approaches. After the convex relaxation, a partial Lagrangian formula, which captures network objective and constraints from the physical layer to application layer, is devised. Then, a cross-layer optimization scheme through a vertical decomposition method (COVD) is proposed, which leads to a novel cross-layer architecture. Through the dual decomposition and primal decomposition, the complex joint optimization problem is decoupled into three subproblems with control parameters flowing back and forth. Although COVD achieves the optimal solution for the joint optimization issue, it incurs high cost in terms of overhead and complexity. Furthermore, we propose a cross-layer optimization design by heuristic algorithm, which reduces the computation complexity by a step-by-step division approach. Finally, simulation results demonstrate the efficiency of the proposed schemes. Complexity analysis and violation of the current protocols are also provided.