We are in the age where mobile devices have become an indispensable part of our lives. They bringrnapplications and services anytime and anywhere. The applications and services rely, mainly, on ourrncellular data plans. As the cellular infrastructure is more reliable and has better coverage, itrnshoulders the majority of user traffic globally. Due to that it is getting overburdened and is forcingrncustomers to go on complaining about poor performances. Though there are many alternativernsolutions for service providers to choose from, mobile traffic offloading is regarded as affordablernand better as a short term solution due to many reasons. Less or no modification to existingrninfrastructure, availability of the required enabling technologies in existing mobile devices, and pricernworthiness of the initial investment and operational cost are among the reasons that appeal to bothrnoperators and customers.rnTherefore, in this research work a feature-rich interworking architecture is proposed for the purposernof mobile traffic offloading. It is fully-integrated, hybrid and Quality of Service -aware. It integratesrnthree different wireless technologies, namely, cellular, wireless local area network and device-todevicerncommunication. Moreover, it has multiple links between the cellular and the other twornnetworks: very tight and loose couplings. In addition, it supports both delay tolerant and delayrnsensitive traffic types. On top of that, a traffic offloading algorithm is devised that can monitor therninstantaneous network conditions and decide through which specific path to en-route the packetsrnwhile fulfilling the Quality of Service demand of the initiated data traffic.rnThe implementation is done using the NS3 simulation tool followed by performance evaluationrnusing six key performance metrics. Packet loss ratio, end-to-end delay, jitter, packet delivery ratio,rnoffloading ratio and handover delay are the applied metrics.rnThe results obtained are very much promising and encouraging. For instance, it is found that up torn90% of delay tolerant traffic and above 50% of delay sensitive traffic can be offloaded from therncellular network towards the wireless local area and device-to-device connections.rnThough the results obtained may not be conclusive the various performance indicators demonstratedrnthe huge potential of such a feature-rich interworking architecture with many possible futurernimprovements.