Mobile healthcare (mHealth) systems integrate patient networks with medical infrastructures to facilitate remote diagnosis of medical conditions reliably and effectively. The rising number of chronic disease and aging patients, mandating in-patient treatment for constant monitoring is posing significant challenges towards high cost healthcare. On the other hand, the proliferation of sensor and smartphone technologies is creating opportunities towards out-patient self-treatment, through leveraging patient networks with cost-effective sensors and mobile devices for smart monitoring and integration with the historical data for effective treatment. In a previous research project, we have focused on patient network and developed a novel framework for wireless Body Area Sensor Networks (BASNs), integrating scalable signal processing based on compressive sensing and wavelet techniques of vital signals such as Electrocardiogram (ECG), Electroencephalogram (EEG), with power-efficient cross-layer communications. However, managing individual patient monitoring networks together with the historical medical infrastructure for reliable and effective integration, remains to be a challenge due to scalability, service disruptions, and data integrity, which calls for new solutions for effective treatment. The goal of the proposed research is to develop a generalized mHealth architecture for reliable, and effective patient monitoring and medical data management, leveraging sensors and smartphone technologies for connecting patient networks with medical infrastructure to facilitate remote patient treatment. The architecture supports heterogeneous modalities such as EEG, ECG, and medical imaging (e.g. MRI and CT X-ray), etc., with diverse Quality of Service (QoS) requirements, and provides innovative solutions for scalable network architecture, effective signal processing, reliable communications, and multimodal analysis for accurate medical diagnoses. To achieve the research goal, we plan to pursue the following research objectives: 1) develop a scalable network architecture integrating patient networks using heterogeneous vital signs such as EEG, ECG, and more, with the medical infrastructure storing historical medical data, to allow for multimodal data correlation and effective diagnosis. 2) Develop new signal processing techniques leveraging compressive sensing for effective processing, energy-efficient communication, and accurate reconstruction of medical data and vital signs. 3) Develop new communication techniques to provide energy-efficient, and delay-efficient transmission of vital signs leveraging cross-layer design and optimization. 4) Develop testbeds to demonstrate the robustness and effectiveness of the proposed signal processing and communication techniques for multimodal data correlation, and remote treatment of medical conditions such as seizure detection & localization, and detection of mental tasks for real-time brain computer interface systems.