| In recent years, the cellular telephone has became one of the requisite personal set. To make the mobile services more attractive, cellular telephones have been equipped with on many new functions beyond just making calls. Mobile TV is the latest and perhaps boldest new capability. The multimedia and mobile trend of the service on terminal foreshows there will be a great market demand on the mobile TV service.Funded by the Hi-Tech R&D Program of China named"Study of the architecture characteristic and the cooperation manner of the next generation RANs"and National Nature Science Foundation of China named"Study on some key technologies on the seamless integration of all IP heterogeneous networks", this thesis studies the on-demand mobile TV service on the heterogeneous networks, including the DVB-H broadcast network, the cellular network and the Wireless LAN. This mainly consists of the following issues:1. The current solutions on the mobile TV service are summarized. The strong point and weak point of each kind of solution are analyzed in detail. Taken advantage of the strongpoint of these solutions, an architecture is proposed in which the multicast on-demand service could be realized by the heterogeneous networks.2. One kind of system block diagram and the chart flow are designed to realize the on-demand mobile TV service by the broadcast network and the cellular network. Also, the main signaling between these networks is schemed out. One kind of system block diagram and the chart flow with the WLAN assistance are further put forward. 3. One kind of modulation recognition approach is studied to identify the various modulation types in the heterogeneous environment. The recognition vectors of the decision-theoretic approach and the cumulant-based approach are combined, and the reference points are trained in RBF neural network to recognize most of the modulation types which may appear in DVB-H, WCDMA and WLAN.4. Queuing theory is applied to analyze the mobile streaming approach and the fixed-group-size multicast approach proposed in this thesis. The Markov birth-death models of these approaches are established on the situation of fixed channel assignment. The situations of unlimited request arrival without considering the impatience of the users and that of limited waiting seats with considering the impatience have been concerned. When adaptive channel assignment is considered, multi-dimension Markov quasi-birth-death model is adopted to analyze this complex queuing problem.5. Monte-Carlo method is adopted to simulate the queuing situation and analyze the system characteristic. The initial value of the multicast group size is determined by the system profit equation. Then the optimal group size is selected according to the simulation results.6. The assistance effect with WLAN is also studied by numerical simulation. The simple direct assistance effect is shown in this study. The smart assistance algorithm is designed. The users on not-so-hot contents are chosen and served by WLAN.Among the works mentioned above, item 2 to 6 embody the innovations research of this paper. Some valuable conclusion can be drawn from the above work:1. The multicast approaches proposed in this thesis can gain much better performance compared with the mobile streaming approach. Furthermore, the adaptive group size approach is even better. To achieve higher system profit, when requests become denser, the group size of the fixed group size approach should be set larger; while the serve threshold of the adaptive group size approach should be set smaller to trigger the service as soon as possible. 2. The direct WLAN assistance approach can improve the overall performance of the whole system. However, the performance declines for the condition with the users served in DVB-H system. In the WLAN smart assistance algorithm, the hot content is judged by comparison of the estimated group forming time to the average waiting time of the users. If the estimated group forming time is larger, this content is a not-so-hot content; this user can be served by WLAN if the WLAN is available. Otherwise the user will still be queued in DVB-H even if the WLAN is available. The simulation results show that even the users served in DVB-H system can gain better performance with this smart assistance approach,.3. The combined parameter modulation recognition approach can recognize high order modulated signals and is robust on the Gaussian white noise compared with the decision-theoretic approach; at the same time, the recognition accuracy rate is enhanced compared with the cumulant-based approach. The modulation recognition approach can recognize most of the frequently used modulation types in DVB-H, WCDMA and 802.11a/b/g. Moreover, this approach can also be used on other modulation recognition stage.In conclusion, this thesis tries to study on the realization of the on-demand mobile TV service in heterogeneous networks full-scaled, including the system block diagram and flowchart, the modulation recognition method which should be used on the signal detection when cooperating between networks, the queue model of the on-demand approaches, and the simulation analysis of the on-demand approaches. The designed system structure and the deliberately chosen parameters will be helpful when this service is put in practice.
|
PDF Full Text Request