Essay On Jamming

The ideal jammer
From the (potential) victim's point of view, it is important to know how an ideal jammer would behave like. Therefore, in this section, a behaviour model of an optimally acting attacker will be developed. Such a model will later help to design more reliable security systems.
In, an ideal jammer is characterized as follows:
He or she should be highly energy efficient. So while attacking, a jammer should consume power as low as possible.
He or she should act stealthy. This means that the probability of detection should also be as low as possible, preferably close to 0%.
He or she should disrupt communications to the maximum possible extent.
He or she should be resistant to various anti-jamming techniques such as jamming mitigation or prevention attacks. In particular, a jammer should forbid signal processing techniques which overcome the attack.

Essentially, it requires to discriminate between legitimate and adversarial causes of poor connectivity: Legitimate scenarios for poor connectivity may be congestion or device failures, while jamming of course belongs to the adversarial causes.
A common strategy in jamming detection is to select certain metrics which are influenced during jamming, manually define thresholds for each of them and finally observe if one or more of these thresholds are exceeded. However, this technique is not unproblematic which the following questions will show:
How many metrics have to be selected in order to detect jamming reliably?
Which threshold values should be assigned to the individual metrics?
In particular, the first question captures the problem that one tends to use a higher number of metrics as usual when dealing with harder-to-detect jammers like reactive jammers. Of course, more data does not necessarily imply higher precision in evaluation. So the answers to these questions are highly