Integrated Soft Media/Channel Image Communications Consumer Opportunities in QoS Control Nikola Rozic,Mladen Russo, Dinko Begusic, Josko Radic University of Split, R. Boskovica bb. 21000 Split, Croatia Abstract -In this paper we propose exploitation of a great amount of redundancy usually communicated at the user level. A well- known soft channel decoding approach used in telecommunications is extended to a “soft media decoding” concept and an integrated media/channel soft decoding scheme has been proposed. Simulation results show the efficiency of the proposed approach for typical image communications over different channels. The main advantage of the proposed integrated media/channel soft decoding scheme is the possibility of repetition (or replication) decoding when different source encoding methods are used. Such methods include different quantization methods, different JPEG qualities or even different coding schemes such as DPCM, JPEG, DWT etc.. Keywords-redundancy; soft decoding; media and channel decoding; I. INTRODUCTION Soft concepts are becoming the most significant feature of the advanced communication and information systems. An important application area of the soft concepts in telecommunication systems is error control and channel coding [1, 2]. Success of the soft approach in telecommuni- cations inspires us to consider the applicability of channel coding methods in real life data communication and information systems. The main motive for this paper is the fact that there is a great amount of redundancy at the user level that is not exploited by the telecommunication system. For example, the user receives N times the same information which corresponds to the repetition (N,1) code. Information may be sent by using different communication systems (GSM, WLAN, Bluetooth, xDSL, ISDN, Web,…), different source coding formats (PCM, ADPCM, JPEG, DWT,…), different noisy channels, different error control schemes (FEC, ARQ, hybrids…), different channel coding schemes (block, convolutional, concatenated..), and different coding rates and gains. Even more, a common situation is using different terminals (Mobiles, PDA, Laptop, Desktop,…) at different locations and in different times (Fig.1). If we assume unreliable data transmission through telecommunication systems, the redundancy at the user level may be used in order to increase the reliability of user’s data. To do this we need to apply “media soft decoding”. This requires that the channel soft decoder delivers not only data but also their reliability. This means that the channel and media decoders have to be integrated. Media decoding must be performed over the real number field just as it is the case with the supposed media encoding. Since media decoding is performed over the source alphabet it can entirely exploit the source redundancy to improve MAP or MS decisions. It is also important to note that in human communications the embedded redundancy Internet Web pages Data IBM Compatible Comm. Tower Television Satellite dish Satellite Minicomputer Plotter Publishing houses data 1 data 2 data 3 data 4 data source Figure 1 Typical public data communications is “gratis”. People use to repeat sentences, data, photos etc. without control. Besides that, if we consider off-line applications, the computing complexity and the delay are not critical issues. All the above-mentioned makes the benefits achieved by using media decoding even more significant. Section 2 describes the proposed integrated media/channel soft decoding scheme. Some basic relationships on soft decoding over the real (integer) number field are presented in Section 3. In Section 4 a soft digital to analog conversion, soft DPCM and soft IDCT are analyzed since are used in the proposed scheme. Simulation results presented in Section 5 demonstrate the usability and efficiency of the proposed approach for typical image communications in real life. Some conclusion remarks are given in Section 6. II. INTEGRATED MEDIA/CHANNEL SOFT DECODING People are used to communicate, directly or via telecommunication system, by using a lot of redundancy. This redundancy appears through multiple data repetition and data aggregation. Simple repetition code RC(N,1) or single parity code SPC (N,N-1) are not good codes from the point of view of their application in telecommunications. However, in public communications where ‘transmitter energy’ is ‘gratis’, each code is a good code. Every data, no matter how unreliable is, carries a piece of information that a soft media decoder may extract and use to improve the data delivered to the user. To increase the reliability of data delivered to the recipient a media soft decoding (Fig.2) as an outer process concatenated with an inner soft channel decoding process can be performed. To realize it, the channel soft decoder should deliver to the 0-7803-8785-6/05/$20.00 (C) 2005 IEEE