A multi-functional technology in the wireless sensor networking (WSN) area is tackled in this paper: routing optimization (RO) with scheduling, channel, and power assignment (SCP) in multi-power multi-radio (MPMR) WSNs. The authors use the linear programming optimization method and provide a novel polynomial-time heuristic scheme with the design and description of an efficient distributed routing algorithm based on random walks.

After discussing related literature, the authors introduce a model: a network of sensors equipped with multi-channel, MPMR transmission-level instruments and the weighted energy and delay (WED) objective function. Finding the optimal scheduling time among the concurrent transmission link sets (CTLSs) also would optimize the objective function of the paper, according to the authors. After defining a rough and coarse linear program problem, the authors try to find the optimum sets of CTLS. By considering the non-optimality of the solution because of the exponential growth of the complexity with the increase in system components, they propose new algorithms based on pruning the number of independent CTLSs.

To accomplish the pruning, a polynomial-time heuristic algorithm has been used, which includes a two-phase method: a routing algorithm to find the paths to achieve the max-net-workflow and estimate the traffic load on each path, and the ability to assign available channels to the links of the paths with non-zero traffic load. The authors assume the WSN is an interference-free network and solve another maximal network flow linear programming (LP) equation, allowing them to acquire the optimal routes. The solution is implemented by assigning available channels to the links of the derived routes and scheduling them concurrently. This solution is refined via bypassing the max-flow formulation and exploiting another random-walk distributed algorithm, using local neighbor information. It uses an upgraded version of “flooding,” based on the gathered information from its child neighbors, the neighbors that are connected to other power levels beyond the optimal power level (OPL). Finally, using the received information from the destination, the receiver--after applying the repetitive branch-bound method, based on the WED objective function--stabilizes the good communication paths. This algorithm is extended to a multi-path algorithm to maximize concurrent transmission.

The announced results in terms of performance evaluation are worthy and considerable; however, the method is remarkably different than reality because of the effective factors of wireless communication-related phenomena and their undeniable influence on the operation of the system.