A key challenge to the development of next-generation mobile ad hoc networks is to enable mobile users to access, manipulate and distribute voice, video, data and images with suitable quality of service (QOS). Mobile ad hoc networks are multi-hop in nature, bandwidth limited and are currently designed to support only best effort voice and data communications. Nodes that constitute the wireless network infrastructure are free to move randomly and organize themselves in arbitrary fashions. Thus, the network's wireless topology may be relatively static over periods of time or, more likely, change rapidly in unpredictable ways. Such a network may operate in a standalone fashion, or may be interconnected to intranets or the global Internet via gateway routers. Providing quality of service support to mobile ad hoc networks is the subject of this proposal. There is a critical need to propose, investigate and evaluate quality of service support in mobile ad hoc networks and to study their interconnection to the global Internet. The approach will be to investigate quality of service in mobile ad hoc networks assumed to be interconnected to the global Internet through broadband cellular networks. One of the primary objectives of this proposal is to determine the level of quality of service assurances that can be given to mobile hosts and routers in such an environment. Quality of service support needs to be highly adaptive and responsive to changes in the available resources along the path between two communicating mobile hosts and to network topology changes due to host and router mobility. The researcher propose a research agenda that comprises the following project phases. First, study and define a comprehensive QOS architecture and associated algorithms for maintaining adaptive real-time services in mobile ad hoc networks that can be interconnected to the global Internet through broadband cellular access points. Second, investigate, define, simulate and deploy wireless flow management based on in-band signaling and the notion of mobile soft-state management capable of establishing, maintaining and terminating adaptive real-time flows in mobile ah hoc networks. Third, study, define and analyze new admission control algorithms that allocate bandwidth to adaptive real-time flows based on the notion of utility-fair allocation that supports minimum bandwidth assurances. Next, investigate, design and analyze suitable wireless packet scheduling mechanisms that can support the packet level quality of service while responding to location dependent channel conditions experienced in ad hoc wireless networks. Finally, extend our existing broadband cellular testbed to the ad hoc domain and demonstrate proof-of-concept of wireless flow management, admission control and wireless packet scheduling in an experimental setting. Throughout the course of the project we will validate these algorithms using a combination of analytical, simulation and experimental techniques.