Multi-constrained quality-of-service routing (QoSR) is to find a feasible path that satisfies multiple constraints simultaneously, as an NPC problem, which is also a big challenge for wireless mesh networks. In this paper, we propose a novel Routing Scheme based on Mean Field Annealing (MFA_RS) to solve this problem. MFA_RS first uses a function of the wireless link’s delay and capacity as the cost function, both of which are QoS parameters, and then seeks to find a feasible path by mean field annealing. Because MFA_RS uses a set of deterministic equations to replace the stochastic process in Simulated Annealing (SA), and uses saddle point approximation in the calculation of the stationary probability distribution at equilibrium, thus the convergence time is much less than the algorithm based on SA. Simulation results demonstrate that MFA_RS is an effective algorithm and is very fit for wireless mesh networks where the topology may be changing constantly.

An alternative blind adaptive multiuser detection is investigated based on modified constrained constant modulus (CM) criterion. It is shown that the performance of a CM-based receiver is limited by the received power of the desired user. In this paper, we show that the limitation can be avoided using the noncanonical constraint CM criterion and that in the presence of channel noise the modified CM criterion function is strictly convex by properly selecting some constant. With analyzing the extrema of the cost function, we point out how to select the constant. Moreover, a simple stochastic gradient algorithm for implementing our scheme is presented, and the convergence properties of the algorithm are analyzed. Simulation examples are given to demonstrate the performance of the proposed scheme.

This paper presents a novel learning algorithm for the approximate reasoning-based Fuzzy Adaptive Mapping Network (FAMN). In the proposed algorithm, we use a serial genetic algorithm to adjust the membership function of the rules. In particular, unlike other adaptive methods, the learning is achieved by incorporating the idea of multiple resolution. The tuning is first implemented with few rules at the coarsest resolution of input/output variables, then with many rules at the higher resolution until the training precision required is obtained. The 2D sine function net is constructed as an illustrative example.