<p>Intrusion signals in mobile communication networks are often disguised as normal communication signals to attack. This impersonation attack is highly covert across the full link, because the confusion of signal features at cross-node of different links is more complicated. This makes it difficult to be accurately recognized by traditional detection method with single point. Meantime, the risk of data breach caused by the attack will generate a ripple effect due to complexity and user mobility in mobile communication network, and will heavily increases the danger of data breach. For this reason, this paper proposes a full link security defense algorithm against malicious intrusion signals in mobile communication networks based on data-driven technique. This algorithm uses the support vector machine (SVM) technology to cons truct an identification model against the malicious intrusion signal of the full link and introduces the firefly algorithm to optimize the support vector parameters of the model to ensure the accuracy of the model in identifying the malicious intrusion signal. In addition, this algorithm uses a network full link security defense model based on dynamic camouflage technology to dynamically simulate any element of the full link in the mobile communication network, and at the same time constructs heterogeneous executors to distribute the results of the malicious intrusion signal to each selected heterogeneous executor. Experimental results show that the proposed algorithm can accurately identify different types of malicious intrusion type signal samples, so that the interception rate of the intrusion defense system against malicious intrusion signals is greater than 99%, and the important data loss rate is less than 1%.</p>