In an advanced wireless network, trust is desirable for all routing protocols to secure data transmission. An enormous volume of important information communicates over the wireless network using trusted dynamic routing protocol, which is the enhancement of the DSR (Dynamic Source Routing) protocol to improve trust. Previously fuzzy logic, genetic algorithm, neural network has been used to modify DSR and good result has been obtained in few performance indicators and parameters. In this work an SVM based trusted DSR have been developed and better results have been presented. This new novel on demand trust based routing protocol for MANET is termed as Support vector machine based Trusted Dynamic Source Routing protocol, performance of STDSR has been improved in term of the detection ratio (%) at different mobility and no. of malicious node variation.

An ad-hoc network is a set of mobile nodes in which it is required that each node performs cooperatively and a node is called cooperative when it transfers data correctly to another node in a wireless network. But due to openness in ad-hoc network, it is vulnerable to various kinds of attack from malicious nodes. Various routing protocol recently have been projected to perform secure routing. For the identification of malicious nodes in mobile ad-hoc network trust based reactive routing protocol are typically used and consequently achieved results are far better. This paper surveys totally different trusted Dynamic source routing protocol, and it is analyzed that the dynamic source routing protocol performs better, whenever the thought of trust being placed in the straightforward dynamic source routing protocol.