The wireless ad-hoc networks consist of autonomous nodes. Data transmission in these networks largly depends on the cooperativity among the nodes. Without it neither routs can be estabilished nor data can be delivered. The main problem is that autonomous nodes exhibit selfish behavior because they want to optimize their performance only. An incentive scheme is needed to stimulate the cooperation among selfish nodes in order to make the network working properly. In this work a credit-based 3-D Markov chain model is introduced to analyze the packet dropping probability with given bandwidth, buffer space and maxumum credit. Based on the Markov chain model the bandwidth distribution is optimized such that the dropping probability of own packets is minimized. The credit based scheme governs the nodal behavior and stimulate the cooperation among nodes, but it can’t optimize the traffic load and the network wide stimulation problem because it uses local information only. Accordingly a cooperativity-based dynamic routing procedure is intruduced to balance the traffic load and stimulate the cooperation across the network. Analytical and simulation results show how the scheme works in networks of selfish nodes.