Trade-off between node selection and space diversity for accurate uncooperative localization

We consider Received Signal Strength (RSS) based localization and trade-off between space diversity and node selection approaches in terms of localization accuracy. In LOS/NLOS (Line Of Sight/Non Line Of Sight) mixed environment and uncooperative scenario, we show that when shadowing variance increases, we move from a first region where diversity is preferable to selectivity (use of all RSS leads to better accuracy) to a second region where selective aspect leads to better performance (use of higher RSS). Thus, the idea of computing an optimum threshold σoptimal2 on the shadowing variance (respectively γoptimal on the RSS values) which represent the edge between these two regions. This threshold, γoptimal, will indeed help to trade-off between using all nodes or only a subset of them for more accurate localization. The closed form expression for the failure rate, defined as the event when all RSS measures are below a threshold value, (RSS < γoptimal) is first derived. Then, we proposed a scheme for optimizing the choice between node selection and diversity approaches.