Body Area Sensor Networks are low cost sensor networks, very often wireless, that are designed to sense physiological parameters, such as, e.g., heart rate and blood pressure, and that allow easy access to user´s critical and non-critical data. Among the wide range of application fields which can benefit from the use of BASNs, the ones related to health care are particularly appealing, since the capacity of continued monitoring, on-line data processing and in-time communication with the user and/or with health-care professionals promises both better quality of life to persons affected by possibly serious illnesses and preventive health-care to a wider set of potential users (in particularly, to an elderly population). BASNs may also represent a fundamental tool for athletes and fitness enthusiast seeking to improve their performances while not exceeding their physical limitations.Body Area Sensor Networks share most design challenges with other embedded systems, in particular with wireless sensor networks, but are characterized by specific aspects as well. BASNs, for instance, have to be non intrusive, must rely on smaller nodes compared to other types of sensor networks, and their energy supply is even more limited than usual (batteries will be very small, energy scavenging provides at present fairly limited energy supply). Another challenge, due to possible application areas such as personal health care and fitness, derives from the sensitive nature of collected, processed and transmitted data, requiring the use of strong security mechanism. Furthermore, BASNs are easily deployed in environments characterized by relevant noise, which in turn impacts not only on system security but also on data (and communication) robustness.In this project we address the problem of security for BASNs, in the light of the new possibilities and challenges provided by novel technological libraries. In particular we aim at providing BASNs with strong cryptographic primitives and with robustness against physical attacks, and at evaluating the effect of such design decisions on the communication protocol. To achieve our goal, we propose applying an approach that, rather than adapting or redesigning security algorithms to meet the needs of limited-resource devices, takes advantage of the novel technological libraries to develop novel devices supporting standard algorithms. Furthermore, our methodology aims at considering all the design variables since the beginning of the design process, evaluating the effects that each optimization step in one direction has on the other parameters, in particular with reference to adoption of new, very low-power technologies.