Abstract: This paper describes a novel research approach for multi-standard network convergence, integrating multiple broadcast networks for high-speed download with communications networks for the uplink and feedback channel. In particular, we study the integration of DVB-H and the Chinese digital television standard (DTMB) within a unified middleware framework.
Introduction
Mobile users increasingly demand high-bandwidth multimedia services "anywhere, anytime," yet current infrastructure remains fragmented. Cellular networks (2G/3G) provide the necessary uplink for interaction but lack the downlink capacity for mass-market high-definition video delivery. Conversely, digital broadcast networks offer massive downlink throughput but are strictly one-way and vary significantly by region—for instance, DVB-H is prevalent in Europe, while DTMB is the standard in China.
The MING-T project addresses this challenge of network convergence by creating a software and hardware framework that allows a mobile device to switch between these standards transparently. By treating disparate technologies as a single converged network, the system enables seamless high-speed mobile multimedia delivery and "vertical handovers" between different network types.
Layered Middleware Architecture
The core contribution of MING-T is a layered middleware architecture designed to abstract underlying hardware complexities from user applications. This approach ensures that the specific transmission technology—whether it be DVB-H, DTMB, or UMTS—is transparent to the upper software layers.
Key components of this architecture include:
- Convergence Sublayers: Positioned above hardware drivers, these layers translate raw bitstreams into standard IP packets, using IP as the "universal glue" for diverse hardware.
- Unified Electronic Service Guide (ESG): Instead of separate menus for different networks, the system aggregates service information from all available broadcast and interactive channels into a single, unified interface.
- Vertical Handover Algorithms: The system monitors signal strength and bit-error rates to automatically switch data streams between standards (e.g., moving from a DVB-H coverage area into a DTMB area) without interrupting the user experience.
- Scalable Video Coding (H.264 SVC): The framework utilizes layered video encoding. A robust "base layer" ensures basic reception in poor conditions, while "enhancement layers" add detail when high-bandwidth broadcast channels are available.
Evaluation and Testbed Implementation
As a systems-oriented project, MING-T was validated through eight specific "Service Scenarios" rather than traditional static datasets. These scenarios included unified service discovery, access to multiple networks, protected content (pay-per-view via cellular return channel), video-on-demand, and seamless handover between broadcast standards.
Evaluation was conducted on a physical testbed at Tsinghua University in Beijing. This environment utilized actual DTMB and DVB-H transmitters alongside a campus-wide WLAN and cellular infrastructure. Since single-chip dual-mode receivers did not exist at the time, the terminal was prototyped using a notebook computer equipped with multiple specialized receiver cards, proving the middleware's ability to handle concurrent hardware streams.
Key Insights and Results
The project demonstrated that IP-based convergence is the most viable path for integrating diverse hardware. A significant technical hurdle addressed was the difference between China’s DTMB (which uses TDS-OFDM for faster synchronization) and Europe’s DVB (which uses CP-COFDM). The middleware successfully bridged these fundamentally different physical layer behaviors.
Furthermore, the integration of a cellular "return channel" transformed broadcast from a "dumb" one-way pipe into a reliable interactive system. By using an optimized Negative Acknowledgment (NAK) protocol that bundles requests and uses bitwise XOR-sums, the system can repair data for multiple users simultaneously, preventing feedback "implosion."
Results showed successful bandwidth integration ranging from 10 kbps (GSM) to 30 Mbps (DVB-T/DTMB). While voice calls tolerate 100ms dropouts, video is more sensitive; the MING-T framework achieved near-"unnoticed" handovers through the coordination of middleware and scalable coding.
Limitations and Future Work
While the prototype proved the concept, several challenges remain for commercial adoption. Hardware miniaturization is a primary concern, as moving from a notebook-sized setup to a handheld device requires integrated dual-mode silicon. Additionally, while technical handovers are possible, the business logic of roaming—including complex billing and authentication across different international operators—remains an open area for future research.
Summary updated