Video streaming is an increasingly popular Internet application. However, despite its popularity, real-time video streaming still remains a challenge in many scenarios. Limited home broadband bandwidth and mobile phone 3G bandwidth means many users stream videos at low quality and compromise on their user experience. To overcome this problem, we propose CStream, a system that aggregates bandwidth from multiple co-operating users in a neighborhood environment for better video streaming. CStream exploits the fact that wireless devices have multiple network interfaces and connects co-operating users with a wireless ad-hoc network to aggregate their unused downlink Internet bandwidth to improve video quality. CStream dynamically generates a streaming plan to stream a single video using multiple connections and continuously adapts to changes in the neighborhood and variations in the available bandwidth. We have built CStream and evaluated it on a controlled test-bed of computers with various performance measures. The results show linear increase in throughput and improved video streaming quality as the number of cooperating users in a neighborhood increase.

Smartphones now account for the majority of all cell phones in use today [23]. Ubiquitous Internet access is a valuable feature offered by these devices and the vast majority of smartphone applications make use of the Internet in one way or another. However, the bandwidth offered by these cellular networks is often much lower than we typically experience on our standard home networks, leading to a less-than-optimal user experience. This makes it very challenging and frustrating to access certain types of web content such as video streaming, large file downloads, loading large webpages, etc. Given that most modern smartphones are multi-homed and are capable of ac- cessing multiple networks simultaneously, this thesis attempts to utilize all available network interfaces in order to achieve the aggregated bandwidth of each to improve the overall network performance of the phone. To do so, I implement a bandwidth aggregation system for iOS that combines the bandwidths of multiple devices located within close proximity of each other. Deployed on up to three devices, speedups of up to 1.82x were achieved for downloading a single, 10mb file. Webpage loading saw speedups of up to 1.55x.

Despite the growing number of mobile devices equipped with multiple networking interfaces, they are not using multiple available networks in parallel. The simple network selection techniques only allow for single network to be used at a time and switching between different networks interrupts all existing connections. This work presents system that improves network connectivity in presence of multiple network adapters, not only through better network handovers, smarter network selection and failure detection, but also through increased bandwidth offered to the device over aggregated channels. The biggest challenge such a system has to face is the heterogeneity of networks in mobile environment. Different wireless technologies, and even different networks of the same type offer inconsistent link parameters like available bandwidth, latency or packet loss. The wireless nature of these networks also means, that most of the parameters fluctuate in unpredictable way. Given the intended practicality of designed system, all that complexity has to be hidden from both client-side applications and from the remote servers. These factors combined make the task of designing and implementing an efficient solution difficult. The system incorporates client-side software, as well as network proxy that assists in splitting data traffic, tunnelling it over a number of available network interfaces, and reassembling it on the remote side. These operations are transparent to both applications running on the client, as well as any network servers those applications communicate with. This property allows the system to meet one of the most important requirements, which is the practicality of the solution, and being able to deploy it in real life scenarios, using network protocols available today and on existing devices. This work also studies the most critical cost associated with increased data processing and parallel interface usage - the increase in energy usage, which needs to remain within reasonable values for this kind of solution being usable on mobile devices with limited battery life. The properties of designed and deployed system are evaluated using multiple experiments in different scenarios. Collected results confirm that our approach can provide applications with increased bandwidth when multiple networks are available. We also discover that even though per-second energy usage increases when multiple interfaces are used in parallel, the use of multi-interface connectivity can actually reduce the total energy cost associated with performing specific tasks - effectively saving energy.

Despite the growing number of mobile devices equipped with multiple networking interfaces, they are not using multiple available networks in parallel. The simple network selection techniques only allow for single network to be used at a time and switching between different networks interrupts all existing connections. This work presents system that improves network connectivity in presence of multiple network adapters, not only through better network handovers, smarter network selection and failure detection, but also through increased bandwidth offered to the device over aggregated channels. The biggest challenge such a system has to face is the heterogeneity of networks in mobile environment. Different wireless technologies, and even different networks of the same type offer inconsistent link parameters like available bandwidth, latency or packet loss. The wireless nature of these networks also means, that most of the parameters fluctuate in unpredictable way. Given the intended practicality of designed system, all that complexity has to be hidden from both client-side applications and from the remote servers. These factors combined make the task of designing and implementing an efficient solution difficult. The system incorporates client-side software, as well as network proxy that assists in splitting data traffic, tunnelling it over a number of available network interfaces, and reassembling it on the remote side. These operations are transparent to both applications running on the client, as well as any network servers those applications communicate with. This property allows the system to meet one of the most important requirements, which is the practicality of the solution, and being able to deploy it in real life scenarios, using network protocols available today and on existing devices. This work also studies the most critical cost associated with increased data processing and parallel interface usage - the increase in energy usage, which needs to remain within reasonable values for this kind of solution being usable on mobile devices with limited battery life. The properties of designed and deployed system are evaluated using multiple experiments in different scenarios. Collected results confirm that our approach can provide applications with increased bandwidth when multiple networks are available. We also discover that even though per-second energy usage increases when multiple interfaces are used in parallel, the use of multi-interface connectivity can actually reduce the total energy cost associated with performing specific tasks - effectively saving energy.

Smartphones form an emerging mobile computing platform that has hybrid characteristics borrowed from PC and feature phone environments. While maintaining great mobility and portability as feature phones, smartphones offers advanced computation capabilities and network connectivity. Although the smartphone platform can support PC-grade applications, the platform exhibits fundamentally different characteristics from the PC platform. Two important problems arise in the smartphone platform: how to mobilize applications and how to deliver them effectively. Traditional application mobilization involves significant cost in development and typically provides limited functionality of the PC version. Since the mobile applications rely on the embedded wireless interfaces of smartphones for network access, the application performance is impacted by the inferior characteristics of the wireless networks. Our first contribution is super-aggregation, a rapid application delivery protocol that in tandem uses the multiple interfaces intelligently to achieve a performance that is ``better than the sum of throughputs' achievable through each of the interfaces individually. The second contribution is MORPH, a remote computing protocol for heterogeneous devices that transforms the application views on the PC platform into smartphone-friendly views. MORPH virtualizes application views independent of the UI framework used into an abstract representation called virtual view. It allows transformation services to be easily programmed to realize a smartphone friendly view by manipulating the virtual view. The third contribution is the system design of super-aggregation and MORPH that achieve rapid application delivery and mobilization. Both solutions require only software modifications that can be easily deployed to smartphones.