A survey on non-linear optimization problems in wireless sensor networks

Wireless Sensor Networks (WSN) pose several problems in terms of scalability and energy consumption. Many of them have been treated in the literature by means of optimization techniques. Depending on the optimization goal and constraints, a great number of these problems are solved using Linear Programming (LP) techniques whenever possible or the network model is further simplified to fit into this category, albeit accuracy may be degraded. In many other cases, simplifying the network model is not feasible and optimization requires more complex techniques. For these cases, non-linear optimization becomes an interesting alternative. Non-linear optimization is beneficial provided that more accurate results are needed or linear optimization is not achievable or yields unreliable results. This paper surveys recent and significant contributions regarding non-linear optimization problems in WSN, selecting the most relevant objectives to optimize and classify the problems under consideration into several categories. Furthermore, this work provides insights in many important problems requiring non-linear optimization and, finally, discusses current open questions in this area.