<p>With the widespread adoption of mobile cloud computing (MCC) in sectors such as healthcare, transportation, and smart infrastructure, distributed denial of service (DDoS) attacks have become increasingly frequent and sophisticated, posing critical security challenges. Enhancing defense mechanisms against such attacks has thus become essential, particularly in highly dynamic MCC environments. This paper addresses the problem of DDoS attack detection in MCC and considers a defense architecture integrating prevention, detection, and mitigation mechanisms while focusing on the detection component. An adaptive machine learning model based on an evolutionary recurrent self-organizing map (ERSOM), enhanced with K-means clustering, is proposed for detecting botnet-based DDoS attacks. The proposed detection approach supports real-time analysis at the edge and is evaluated under identical experimental conditions against existing techniques. Experimental results demonstrate that the proposed model achieves competitive performance across key evaluation metrics.</p>