Wireless Body Area Network (WBAN) aims to monitor patient's health remotely, by using mini medical sensors that are attached on the human body to collect important data via the wireless network. However, this type of communication is very vulnerable to various types of attacks, poses serious problems to the individual's life who wears the nodes. In this paper, we present a new classification of the most dangerous attacks based on different criteria, which gives us a clear vision of how attacks affect a WBAN system. Moreover, this classification will help us to specify the strength and the weakness of each attack in order to facilitate the development of a new intrusion detection system (IDS). In the second part of this work, we develop a novel IDS for detecting three types of jamming attacks in WBAN. The proposed methodology is based on the network parameters as an indicator to differentiate the normal case from the abnormal case like false alert or attack state. Through simulation analysis that was applied on Castalia platform by using OMNET++ as a simulator, proves that the proposed IDS have a great effect for detecting the presence of jamming attack in the network.