In wired networks, systems are usually optimized to offer the maximum throughput of point-to-point and generally well identified transmissions. In the first widespread wireless networks such as IEEE 802.11, the model of point-to-point communication still applies; the transmissions are between the wireless nodes and the access point, which usually serves as a gateway to the Internet. Yet this model is no longer valid with more recent wireless systems such as Wireless Sensor Networks (WSNs) and Vehicular Ad Hoc NETworks (VANETs). In such networks, communication is between one node and its neighbors and simultaneous transmissions or, in other words spatial reuse, is required to insure good performance. Thus performance is directly linked to the density of successful simultaneous transmissions. Another important remark concerning wireless networks, past and present, is that most of these communications systems use Carrier Sense Multiple Access (CSMA) techniques. Previous studies, such as [1], show that in CSMA networks, the density of successful transmissions is greatly influenced by the carrier sense detection threshold, which is one of the main parameters of CSMA. In this paper, we use a simple stochastic model for CSMA to experiment an adaptive scheme which tunes the carrier sense threshold to the density of network nodes. This model uses a Matern selection process with a random pattern of nodes distributed as a Poisson Point Process (PPP). Each node in the process receives a random mark and the nodes that have the smallest mark in their neighborhood are elected for transmission, mimicking the election process of CSMA. The analysis in this paper indicates that an adaptive technique can be based on the average waiting times of packets. Alternatively, this technique can also be based on the number of neighboring nodes.