No description available.

No description available.

No description available.

No description available.

No description available.

Detta arbete börjar med att presentera en kodningssats gällande linjärkodning över ändliga ringar för kodning av korrelerade diskretaminneslösa källor. Denna sats inkluderar som specialfall motsvarandeuppnåbarhetssatser från Elias och Csiszár gällande linjär kodning överändliga kroppar. Dessutom visas det att för varje uppsättning av ändligakorrelerade diskreta minneslösa källor, så finns alltid en sekvens avlinjära kodare över vissa ändliga icke-kropp-ringar som uppnårdatakompressionsgränsen bestämd av Slepian-Wolf-regionen. Därmed slutervi problemet med linjär kodning över ändlig icke-kropps-ringar föri.i.d. datakomprimering med positiv bekräftelse gällande existens. Vi studerar också kodning av funktioner, där avkodaren är intresseradav att återskapa en diskret mappning av data som genererats av flerakorrelerade i.i.d. källor och som kodats individuellt. Vi föreslårlinjär kodning över ändliga ringar som en alternativ lösning på dettaproblem. Vi visar att linjär kodning över ändliga ringar presterarbättre än sin ändliga-kropp-motsvarighet, liksom dessutomSlepian-Wolf-kodning, i termer av att uppnå bättre kodningshastigheterför kodning av flera diskreta funktioner. För att generalisera ovannämnda genomförbarhetssatser, både gällandedatakompression och funktionskodningsproblemet, till Markov-källor(homogena irreducerbara Markov-källor), så introducerar vi ett nyttkoncept gällande klassificering av typiska sekvenser, benämndSupremus-typiska sekvenser. Den asymptotiska likafördelningsprincipensamt en generaliserad version av typiskhets-hjälpsatsen förSupremus-typiska sekvenser bevisas. Jämfört med traditionell (stark ochsvag) typiskhet, så tillåter Supremus-typiskhet oss att härleda bättretillgängliga verktyg och resultat, som låter oss bevisa att linjärkodning över ringar är överlägsen andra metoder. I motsats härtillmisslyckas argument baserade på den traditionella versionen antingen medatt nå liknande resultat eller så är de härledda resultaten svåra attanalysera på grund av en utmanande utvärdering av entropitakt. För att ytterligare undersöka den grundläggande skillnaden mellantraditionell typiskhet och Supremus-typiskhet och dessutom göra våraresultat än mer allmänt gällande, så betraktar vi ävenasymptotiskt medelvärdesstationära ergodiska källor. Våra resultat visaratt en inducerad transformation med avseende på en ändligt mätbar mängdöver ett rekurrent asymptotiskt medelvärdesstationärt dynamiskt systemmed ett sigma-ändlig sannolikhetsmått är asymptotisktmedelvärdesstationär. Följaktligen så gällerShannon-McMillan-Breiman-teoremet, liksom Shannon-McMillan-teoremet, föralla reducerade processer härledda ur rekurrenta asymptotisktmedelvärdesstationära stokastisk processer. Alltså ser vi att dettraditionella typiskhetkonceptet endast realiserarShannon-McMillan-Breiman-teoremet i ett globalt hänseende, medanSupremus-typiskhet leder till att resultatet håller samtidigt även föralla härledda reducerade sekvenser. / This work first presents a coding theorem on linear coding over finite rings for encoding correlated discrete memoryless sources. This theorem covers corresponding achievability theorems from Elias and Csiszár on linear coding over finite fields as special cases. In addition, it is shown that, for any set of finite correlated discrete memoryless sources, there always exists a sequence of linear encoders over some finite non-field rings which achieves the data compression limit, the Slepian--Wolf region. Hence, the optimality problem regarding linear coding over finite non-field rings for i.i.d. data compression is closed with positive confirmation with respect to existence. We also address the function encoding problem, where the decoder is interested in recovering a discrete function of the data generated and independently encoded by several correlated i.i.d. sources. We propose linear coding over finite rings as an alternative solution to this problem. It is demonstrated that linear coding over finite rings strictly outperforms its field counterpart, as well as the Slepian--Wolf scheme, in terms of achieving better coding rates for encoding many discrete functions. In order to generalise the above achievability theorems, on both the data compression and the function encoding problems, to the Markovian settings (homogeneous irreducible Markov sources), a new concept of typicality for sequences, termed Supremus typical sequences, is introduced. The Asymptotically Equipartition Property and a generalised typicality lemma of Supremus typical sequences are proved. Compared to traditional (strong and weak) typicality, Supremus typicality allows us to derive more accessible tools and results, based on which it is once again proved that linear technique over rings is superior to others. In contrast, corresponding arguments based on the traditional versions either fail to draw similar conclusions or the derived results are often hard to analyse because it is complicated to evaluate entropy rates. To further investigate the fundamental difference between traditional typicality and Supremus typicality and to bring our results to a more universal setting, asymptotically mean stationary ergodic sources, we look into the ergodic properties featured in these two concepts.Our studies prove that an induced transformation with respect to a finite measure set of a recurrent asymptotically mean stationary dynamical system with a sigma-finite measure is asymptotically mean stationary. Consequently, the Shannon-McMillan-Breiman Theorem, as well as the Shannon-McMillan Theorem, holds simultaneously for all reduced processes of any finite-state recurrent asymptotically mean stationary random process.From this, we see that the traditional typicality concept only realises the Shannon-McMillan-Breiman Theorem in the global sequence, while Supremus typicality engraves the simultaneous effects claimed in the previous statement into all reduced sequences as well. / <p>QC 20150225</p>

Speech coding at 64 and 32 Kb/s is well developed and standardized. The next bit rate of interest is at 16 Kb/s. Although. standardization has yet to be made, speech coding at 16 Kb/s is fairly well developed. The existing coders can produce good quality speech at rates as low as about 9.6 Kb/s. At present the major research area is at 8 to 4.8 Kb/s. This work deals first of all with enhancing the quality andkcomplexity of some of the most promising coders at 16 to 9.6 Kb/s as well as proposing new alternative coders. For this purpose coders operating at 16 Kb/s and 12 to 9.6 Kb/s have been grouped together and optimized for their corresponding bit rates. The second part of the work deals with the possibilities of coding the speech signals at lower rates than 9.6 Kb/s. Therefore, coders which produce good quality speech at bit rates 8 to 4.8 Kb/s have been designed and simulated. As well as designing coders to operate at rates below 32 Kb/s. it is very important to test them. Coders operating at 32 Kb/s and above contain only quantization noise and usually have large signal to noise ratios (SNR). For this reason their SNR's may be used for comparison of the coders. However, for the coders operating at 16 Kb/s and below this is not so and hence subjective testing is necessary for true comparison of the coders. The final part of this work deals with the subjective testing of 6 coders, three at 16 Kb/s and the other three at 9.6 Kb/s.

This thesis presents an approach to the implementation of declarative languages on a simple, general purpose concurrent architecture. The safe exploitation of the available concurrency is managed by relatively sophisticated code generation techniques to transform programs into an intermediate concurrent machine code. Compilation techniques are discussed for 1'-HYBRID, a strongly typed applicative language, and for 'c-HYBRID, a concurrent, nondeterministic logic language. An approach is presented for 1'- HYBRID whereby the style of programming influences the concurrency utilised when a program executes. Code transformation techniques are presented which generalise tail-recursion optimisation, allowing many recursive functions to be modelled by perpetual processes. A scheme is also presented to allow parallelism to be increased by the use of local declarations, and constrained by the use of special forms of identity function. In order to preserve determinism in the language, a novel fault handling mechanism is used, whereby exceptions generated at run-time are treated as a special class of values within the language. A description is given of ,C-HYBRID, a dialect of the nondeterministic logic language Concurrent Prolog. The language is embedded within the applicative language 1'-HYBRID, yielding a combined applicative and logic programming language. Various cross-calling techniques are described, including the use of applicative scoping rules to allow local logical assertions. A description is given of a polymorphic typechecking algorithm for logic programs, which allows different instances of clauses to unify objects of different types. The concept of a method is derived to allow unification Information to be passed as an implicit argument to clauses which require it. In addition, the typechecking algorithm permits higher-order objects such as functions to be passed within arguments to clauses. Using Concurrent Prolog's model of concurrency, techniques are described which permit compilation of 'c-HYBRID programs to abstract machine code derived from that used for the applicative language. The use of methods allows polymorphic logic programs to execute without the need for run-time type information in data structures.

No description available.

A performance evaluation of CSTC (Cooperative Space Time Coding) with spatially cor-related fading and imperfect channel estimation in Gaussian as well as impulsive noise is presented. Closed form expressions for the pairwise error probability conditioned on the estimated channel gains are derived by assuming the components of the received vector are independent given the estimated channel gains. An expurgated union bound using the limiting before averaging technique given the estimated channel gains is then obtained. Although this assumption is not strictly valid, simulation results show that the bound is accurate in estimating the diversity order as long the channel estimation is not very poor. It is found that CSTC with block fading channels can reduce the frame error rate (FER) relative to SUSTC (Single User Space Time Coding) with quasi-static fading channels, even when the channel gains for each user are strongly correlated and when the channel estimations are very poor. A decision metric for CSTC with spatially correlated fading, imperfect channel estimation, and impulsive mixture Gaussian noise is derived which yields lower FERs than the Gaussian noise decision metric. Simulation results show that the FER performance of CSTC with mixture Gaussian noise outperforms CSTC with Gaussian noise at low SNR. At high SNR, the FER performance of CSTC with Gaussian noise is better than the FER performance of CSTC with mixture Gaussian noise due to the heavy tail of the mixture Gaussian noise.