The next generation of networks will involve interconnection of heterogeneous individual networks such as WPAN, WLAN, WMAN and Cellular network, adopting the IP as common infrastructural protocol and providing virtually always-connected network. Furthermore, there are many devices which enable easy acquisition and storage of information as pictures, movies, emails, etc. Therefore, the information overload and divergent content’s characteristics make it difficult for users to handle their data in manual way. Consequently, there is a need for personalised automatic services which would enable data exchange across heterogeneous network and devices. To support these personalised services, user centric approaches for data delivery across the heterogeneous network are also required. In this context, this thesis proposes Smart PIN - a novel performance and cost-oriented context-aware Personal Information Network. Smart PIN's architecture is detailed including its network, service and management components. Within the service component, two novel schemes for efficient delivery of context and content data are proposed: Multimedia Data Replication Scheme (MDRS) and Quality-oriented Algorithm for Multiple-source Multimedia Delivery (QAMMD). MDRS supports efficient data accessibility among distributed devices using data replication which is based on a utility function and a minimum data set. QAMMD employs a buffer underflow avoidance scheme for streaming, which achieves high multimedia quality without content adaptation to network conditions. Simulation models for MDRS and QAMMD were built which are based on various heterogeneous network scenarios. Additionally a multiple-source streaming based on QAMMS was implemented as a prototype and tested in an emulated network environment. Comparative tests show that MDRS and QAMMD perform significantly better than other approaches.