Radio frequency (RF) jamming is a denial of service attack targeted at wireless networks. In resource-hungry scenarios with constant traffic demand, jamming can create connectivity problems and seriously affect communication. Therefore, the vulnerabilities of wireless networks must be studied. In this study, we investigate a particular type of RF jamming that exploits the semantics of physical (PHY) and medium access control (MAC) layer protocols. This can be extended to any wireless communication network whose protocol characteristics and operating frequencies are known to the attacker. We propose two efficient jamming techniques: A low-data-rate random jamming and a shot-noise based protocol-aware RF jamming. Both techniques use shot-noise pulses to disrupt ongoing transmission ensuring they are energy efficient, and they significantly reduce the detection probability of the jammer. Further, we derived the tight upper bound on the duration and the number of shot-noise pulses for Wi-Fi, GSM, and WiMax networks. The proposed model takes consider the channel access mechanism employed at the MAC layer, data transmission rate, PHY/MAC layer modulation and channel coding schemes. Moreover, we analyze the effect of different packet sizes on the proposed jamming methodologies. The proposed jamming attack models have been experimentally evaluated for 802.11b networks on an actual testbed environment by transmitting data packets of varying sizes. The achieved results clearly demonstrate a considerable increase in the overall jamming efficiency of the proposed protocol-aware jammer in terms of packet delivery ratio, energy expenditure and detection probabilities over contemporary jamming methods provided in the literature.