A summary of the interoperability issues between the different networks and our middleware approach to solving these issues.

The key objective of the project is to provide the integration and seamless handover between disparate broadcast technologies. Of course, the full solution requires two main components, namely:

• a multi-protocol mobile terminal (e.g smartphone or PDA) on the users' side,

• and corresponding encoders and equipment on the network operators' side.

In both cases, software is required to provide a common set of abstractions and functions that can be re-used between the several broadcast and communication standards. However, the low-level features and functions are not only well-defined but also fixed for each of the systems. Any attempt at interoperability must therefore introduce another layer of software above the existing and fixed functions of the base communication channels.

Therefore, we have already arrived at a very important conclusion. To achieve the goals of interoperability and handover, we have to provide the user with a suitable multi-protocol mobile receiver, and we have to design and implement an additional layer of middleware software that provides the common abstractions for low-level data-transfer, including error-correction and seamless handover. 

The figure below shows a global view of the services required, from a network operators point of view. Note that the different broadcast and communication networks are now summarized as two tiny arrows on the right, while the main range of services is much broader. This should help to explain why we can hope to reach our goal of interoperability: the conflicting low-level encodings are just a small part of the whole system, and the overall user-experience critically depends on a lot of services, most of which can be shared across the broadcast and communication systems.

Multi-standard mobile terminals

Of course, the first step for interoperability between multiple broadcast and communication networks are suitable multi-protocol terminals, e.g. smartphones or PDAs. A comparison of the DVB-H and DTMB standards provides a good example of the issues involved. Both broadcasting technologies roughly occupy 8MHz channel bandwidth, support a hierarchical modulation scheme and IP encapsulation, as well as utilize the well-established OFDM as their baseline technology. However, there are also substantial differences between these two technologies. For example, in the physical layer,

• both the number of sub-carriers and the length of the guard interval are different, resulting in different frame length and data throughput;

• the guard interval used for OFDM symbol is different, leading to different system performance due to synchronization and channel estimation;

• the size of the time interleaver is also different, giving the system a different capability to handle impulsive noises in the time domain;

• the coding scheme is different. DVB-H uses RS as the outer code and convolutional code as the inner code while DMB-T uses the low-density parity check (LDPC) code; and

• rather different implementation complexity. For example, MPE-FEC or the time-slicing scheme to reduce the power consumption at the handheld devices have not been considered for the DTMB standard.

All these differences at the physical layer are a challenge for interoperability on the receiver. However, we are confident that a deep and thorough investigation on the similarities and differences between these two technologies will result in a framework design that enables interoperability. This common interface for both DVB-H and DTMB will be also establish associations and mappings between the two different broadcast transport streams, and therefore provide a solid basis for the mobility management in the converged multi-radio environment..

For mass adoption and deployment of multi-standard mobile terminals, the total system cost is a very importan factor. Hardware requires a high degree of integration to meet the size and power constraints of a mobile terminal – this typically means a single semiconductor device is required. Several companies have already identified the need to produce multi-standard mobile digital television receiver chips to address existing standards such as ISDB-T (Japan), T-DMB (Korea) and DVB-H (Europe). All of these standards use some form of OFDM (Orthogonal Frequency Division Multiplexing) as their basic modulation, and so share some key fatures. However the coding, framing and modulation do differ in many details. The Chinese DTMB standard also uses OFDM modulation, but, as it is a more recent standard, incorporates some more advanced techniques for coding and modulation that enhance its performance compared to the older, existing, standards.

As an example, the Paradiso chipset of project partner Frontier Silicon already provides an integrated multi-protocol receiver solution for DAB, T-DMB, and DVB-H. At the moment, it is not yet clear whether it is possible to also support DTMB with the current chipset; if not, we will use a dual-chipset receiver for the prototypes and field-trials at the end of the project.

Middleware approach and design

As explained above, we expect to develop a special middleware framework that allows to re-use common functions between all of our target systems. One parts of the system runs on the mobile terminals, while other parts of the software are on the network operators' side. The middleware software is used to facilitate interoperation and interaction between broadcast and mobile operator networks. Besides, it can serve interactive applications to support subscriber specific interactive applications and related management services and handle billing (charging) problems.

So far, most state of the art middleware has been designed for use in the realm of static, resource rich environments and hence is not immediately applicable in mobile settings. In the case of the digital broadcasting and mobile convergence, users could move in and out the coverage areas of both broadcast and mobile networks, which will change, for instance, the usage of the networks as well as the available bandwidth, that means mobility results in additional requirements on appropriate middleware support.

We expect the middleware to be highly complex, due to the wide range of functionalities that have to be addressed. For example, consider the following usage model: A mobile terminal is switched on, and then this device immediately identifies the available services via an electronic services guide (ESG). The user orders one of the available services via a secure transaction with the service provider who decides whether a part of the DVB-H downstream will be allocated to provide the service. The decision could be made based on a popularity measure of the requested service, which allows optimized allocation of the digital broadcasting capacity. Content protection is provided such that only those users who have paid for the service have access to it. A negotiated level of quality of service is guaranteed. If the user moves then, the handovers both in the interactive networks (UMTS, WLAN) as well as in the digital broadcasting networks (DVB-H, DMB-T, etc.) are managed seamlessly.

The research issues include:

• Negotiation between different broadcast and mobile operator networks for billing and interactive services. Different kinds of services can be provided based on this, for example, purely subscription, video on demand, usage-based billing w/o network connection, etc.

• Study on the characteristics of the available networks in a predefined area. It includes the analysis of the available bandwidth in both downlink and uplink channels, how they change in different time scales, e.g. hours, days and months, etc.

• Analysis of Interactive application requirements. Popularity measure of the requested services in the coverage areas and load balancing among different networks. The statistics of the requested services will help to provide the services required by the majority of the users while postponing low priority requests. The service can be shared by different networks in order to balance the load among them.

• Handover problems will be trigged when mobile users move in and out of the coverage areas of different networks.

• Managing resources in a fault-tolerant manner, reducing the impact ill-behaved clients have on system stability. The wireless networks suffer from low throughput, high bandwidth variability, high latencies, unreliability, and their susceptibility to sudden connection losses and temporary unavailability. An overall challenge is to strive for suitable techniques keeping applications working in face of all the diverse resulting problems.

• Scalability should be taken into account since there could be several middleware servers, connected with different networks, and serving a potentially high number of concurrent clients.

The aim of this activity is to provide a set of algorithms, to be implemented into the portal part of the middleware, which can optimize the allocation of the capacity of digital broadcasting networks (DVB-H/DMB-T) and the mobile networks. We should therefore focus on a subset of items and demonstrate their feasibility through and ad hoc communication with the middleware client.

The signalling system will also be developed for the interfaces with IP encapsulator, treatment of ESG, etc. The following image sketches a high-level overview of the planned software framework.

Actually, the emergence of mobile computing has created new classes of applications dependent upon the users’ location, context and interaction with their current environment. However, given the constraints of the wireless environment (i.e. weak connection, poor network quality of service and mobile devices with limited resources) developing distributed applications within this domain is a complex task. Therefore, new middleware has emerged to mask such problems from the user. These encompass synchronous (e.g. remote method invocation) and asynchronous (e.g. publish-subscribe and tuple spaces) communication paradigms. However, it is the heterogeneous nature of the solutions that has created a new problem. Only applications and mobile services developed on the same middleware style can interoperate with one another. Last but not least, we mention the bottleneck met when dealing with a large number of users and services (scalability).

Broadcast transport convergence sublayer 

One promising approach for the overall system design relies on IP datacast of the underlying network. We call this the Broadcast Transport Convergence Sublayer, which will be responsible for the handover implementation between DVB-H and DMB-T as illustrated in the terminal logical structure figure (see the next figure). Since the different broadcast standards like DVB-H and DMB-T have different interfaces the Broadcast Transport Convergence Sublayer can provide a consistent interface to the layers above, thus providing mobility support. The key issue here is to find a common utility that can unite and switch the different broadcast transport streams no matter whether they are MPEG-2 or IP. Power consumption, packet loss, and signalling synchronization issues will be considered.

On the network side, we consider an IPDC (internet-protocol datacast layer) to provide a uniform platform for the seamless handover between different broadcast technologies including DVB-H and DTMB. The IPDC brings new characteristics for the converged networks. The benefits are as follows:

• IPDC provides a platform for true convergence of services between DVB-H, DMB-T and mobile telecommunication cellular networks (GPRS/UMTS).

• IPDC allows the coding to be decoupled from the transport layer, thus opening the door to a number of features benefiting handheld mobile terminals including a variety of encoding methods, which only require low power from a decoder (Decoding high bandwidth MPEG-2 encoded streaming video/audio is relatively power consuming).

• IPDC is relatively insensitive to any buffering or delays within the transmission (unlike MPEG-2).

• IPDC is well suited for time-sliced transmission.