K. V. S. Prasad. A calculus of value broadcasts. In Parallel Architectures and Languages Europe, pages 391--402, 1993.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....This means that the sender and the receiver of a message are both blocked until communication occurs. In fact, if the messages are pure synchronization signals carrying no values, there is no distinction whatsoever between input and output. Another style of communication is modelled in CBS [Pra91, Pra93a, Pra93b, Pra93c]. Here, processes communicate by unbuffered broadcast. One aspect of this is that output is an autonomous action, and the sender does not wait for someone to send to. It has no way of knowing how many, or indeed if any, processes pay attention to what it sends. In contrast, the receivers of a ....

....being replaced by those of CBS, an alternative formulation of bisimulation with strength as a parameter is used. The modified proof system is called TAVB . There are several formulations of CBS to choose from. This thesis uses a typed version of Patterned CBS. The syntax is described briefly in [Pra93a], but that paper concentrates on a special case of Patterned CBS, and is furthermore untyped. The type system used here follows [Pra93b] The semantics used is essentially that of [Pra93a] but modified to separate evaluation and communication rules as in the latter part of [Pra93b] When used in ....

[Article contains additional citation context not shown here]

K. V. S. Prasad, A Calculus of Value Broadcasts, in proceedings PARLE'93, LNCS 694, pp. 391--402, Springer Verlag, 1993.

....of process calculi, and of other theories as well. CCS has due to its simple and elegant mathematical framework proven to be an appropriate basis for several timed [HR90, MT90, NS91, W90, W91, W91b] and probabilistic formalisms [JL91, LS88] and a new calculus based on the broadcast paradigm [Pr91, Pr93v, Pr93p]. One could well say that a process calculus can favourably be used as a test bench for experiments with various adaptations, not only as a fixed formalism. An interesting and important question is the relationships between a calculus and the extensions which are based on it. In some cases the ....

....calculi like CCS appeared to be inadequate to model broadcast satisfactorily due to the disparity in communication models. Moreover the traditional handshake has been found difficult to give distributed implementation. The new calculus CBS (Calculus of Broadcasting Systems) addresses these issues [Pr91, Pr93v, Pr93p]. It derives most of its characteristics from CCS, but instead of handshake, communication is now modelled by broadcast. The sending of a broadcast is an autonomous and atomic event. All processes working concurrently with the sender receive the message and can either accept it or ignore it at ....

[Article contains additional citation context not shown here]

Prasad, K.V.S., A Calculus of Value Broadcasts, in proceedings from PARLE'93, LNCS 694, Springer Verlag, 1993.

....Properties of the Translation 12 7. Discussion 13 7.1. What is the Relation to CCS 13 8. Related and Future Work 13 Acknowledgements 14 References 14 Appendix: Proofs from sections 4 6 15 INTERPRETING BROADCAST COMMUNICATION IN SCCS 3 1. Introduction CBS (Calculus of Broadcasting Systems) [Pr91, Pr92] is a process calculus in the style of Robin Milner s CCS [Mi89] It has been developed during the last three years by K.V.S. Prasad at Chalmers University of Technology. The main difference from other traditional process calculi, like CCS for example, is that the usual handshake communication ....

....are omitted from the text and included in the appendix. The last two sections draw conclusions and points to related work. 2. CBS Summary In this section we present a variant of CBS without value passing [Pr91] and with the scoping and renaming operators replaced by the new renaming operator of [Pr92]. We define the following sets ranged over by the variables indicated in parenthesis. 4 U. HOLMER Operator Transmit Read Discard 0 0 : Gamma 0 Monitor a p a Gamma p a p : Gamma a p ( 6= a) Transmit p Gamma p p : Gamma p Choice 1 p Gamma p 0 p q Gamma ....

Prasad, K.V.S., A Calculus of value broadcasts, PMG report 69, Dept. of Comp. Sci., Chalmers Univ. och Techn., Sweden, 1992.

.... [ p 0 OEp [ OEp 0 1 There is a symmetric rule for q. Table 1: The operational semantics of CBS 2 CBS summary In this section we present a variant of CBS with channels but without value passing [7] and with the scoping and renaming operators replaced by the new translation operator of [8]. Details on different versions of CBS may be found in [7] 8] and [9] We define the following sets ranged over by the variables indicated in parenthesis. A set of names: Delta = fa; b; c; g. A set of messages: Delta = Delta [ f g ( Let ] range over f ; g, over f ; g, ....

.... Table 1: The operational semantics of CBS 2 CBS summary In this section we present a variant of CBS with channels but without value passing [7] and with the scoping and renaming operators replaced by the new translation operator of [8] Details on different versions of CBS may be found in [7] [8] and [9] We define the following sets ranged over by the variables indicated in parenthesis. A set of names: Delta = fa; b; c; g. A set of messages: Delta = Delta [ f g ( Let ] range over f ; g, over f ; g, over f ; g, and for S Delta let S] fa]ja 2 Sg. Let A = ....

[Article contains additional citation context not shown here]

K. V. S. Prasad. A Calculus of Value Broadcasts. In LNCS 694. Springer Verlag, 1993.

....a finitary proof system where judgements are relative to properties of the data domain. 1 Introduction The broadcast calculus, CBS, is a value passing process calculus where process intercommunication is achieved by the broadcasting of values. The calculus has been developed in series of papers [8, 9, 12] and a subset has been implemented as an extension to Lazy ML, 10] Here we are concerned with the development of a sound semantic theory for the Broadcast Calculus, in particular the provision of an equational theory and proof system for establishing process identities. Part of the problem is to ....

....rules depend on deductions which can be made in an independent proof system for the data domain. We now outline the rest of the paper. The syntax and operational semantics of the particular version of CBS that we study is presented in Section 2. The calculus is essentially the calculus CBS of [9], augmented with pattern matching on inputs. The input prefix x T of CBS is replaced with x 2 S T where S can be any subset of values. We then define strong barbed bisimulation equivalence for CBS and we characterise the largest congruence contained within it. This we call noisy bisimulation ....

[Article contains additional citation context not shown here]

K.V.S. Prasad. A calculus of value broadcasts. Technical report, Dept. of Computer Science, Chalmers, 1992.

....is therefore one to many and transmission proceeds even if there is no other process listening for a value; transmission is an unblocked action. Moreover, as with most process calculi, this communication is deemed to occur instantaneously. The calculus has been developed in a series of papers [16, 17, 18] and has been implemented as an extension to Lazy ML, 18] Here we are concerned with the development of a semantic theory for the Broadcast Calculus, in particular the provision of an equational theory and proof system for establishing process identities. Work to this effect has been ....

....implemented as an extension to Lazy ML, 18] Here we are concerned with the development of a semantic theory for the Broadcast Calculus, in particular the provision of an equational theory and proof system for establishing process identities. Work to this effect has been investigated by Prasad in [17] where he defines both strong and weak bisimulation equivalence and provides a complete axiomatisation of the strong case for a pure version of the calculus. The problem of finding a complete axiomatisation This work has been supported by the ESPRIT BRA CONCUR2 project, the EPSRC grant ....

[Article contains additional citation context not shown here]

K.V.S. Prasad. A calculus of value broadcasts. Technical report, Dept. of Computer Science, Chalmers, 1992.

....index in the sequence. Two ill ordered values will exchange place by exchanging their index: a; i) b; j) Gamma (a; j) b; i) if a b and i j Greatest common denominator (GCD) of a multiset of numbers: a; b Gamma a; b Gamma a; if 1 a b (1) 3 Introduction to CBS Processes in CBS [Pra93a, Pra93b, Pra95, Wei95] are built from the basic operations v , output (broadcast) the value v , and S , input any value v from the set S . The operations used for translating Gamma programs are described as follows: ffl o, called the empty (or dead ) process, does nothing. ffl The process v Delta P can ....

K. V. S. Prasad. A calculus of value broadcasts. In PARLE'93, volume 694 of LNCS. Springer Verlag, June 1993.

....processes keep track of time. Readership Readers are assumed to be familiar with CCS [Mil89] Familiarity with some timed process calculus and with [Pra93b] would help, but is not strictly necessary. More discussion of the CBS model of computation, particularly of autonomy, can be found in [Pra93a, Pra91]. 2 The Syntax and Semantics of TCBS A TCBS process of type P roc ff says or hears values of type ff and evolves to other processes of type P roc ff. The actions of saying and hearing v: ff are written v 00 and v 00 . Let ffi , ffi 0 range over T , the time domain. This can be taken to ....

....list is a set, and justify the sum notation. In Law 2, P is the set of all processes. Proposition12 Timed laws. 1. ffiP ) v P=v 2. ffi ( f) f 3. ffi (ffi 0 P ) ffi ffi 0 ) P 4. ffi (P j Q) ffi P j ffi Q 5. ffi (OEP ) OE (ffiP ) Expansion theorems CBS , the calculus of [Pra93a], is an untimed calculus with and prefixes, 0 and as primitive constructors. The equations 1(a) and 1(b) of Proposition 11 together with the expansion theorem constitute a complete axiom system for finite processes of CBS . This result has not yet been adapted to the CBS of [Pra93b] ....

K. V. S. Prasad. A calculus of value broadcasts. In PARLE'93, June 1993. Springer Verlag LNCS 694.

....to throw away a lot of communication bandwidth. This paper suggests it is also throwing away much else. Funding: From the Swedish Government agency TFR, and from the Esprit Basic Research Action CONCUR2 A sustained study applying process calculus techniques to broadcast communication ([Pra91, Pra93a, Pra93b, Pra94], also [Pra95, Jon93, Hol93, Pet94] has yielded what promises to be an elegant calculus of broadcasting systems, CBS, and some elegant programs. CBS is fashioned after CCS [Mil89] but the change from handshake to broadcast communication has far reaching consequences both for programming and for ....

....reports the stable aspects of CBS as of now, including some material not previously presented. It concentrates on basic theory, on programming [Pra93b] and on priority [Pra94] The current version of CBS is a subcalculus of that of [Pra93b] and is significantly different from those of [Pra91] and [Pra93a]. Organisation of the paper. A sequential pass should encounter few forward references. Ignoring the occasional outside reference, many of Sections 3 to 11 can be read immediately after Section 2 (concepts) and any of Sections 13 to 15 after Section 12 (adding priorities to CBS) Sections 2 and ....

[Article contains additional citation context not shown here]

K. V. S. Prasad. A calculus of value broadcasts. In PARLE, June 1993. Springer Verlag LNCS 694.

....report on ongoing work: Higher Order Broadcasting Systems (HOBS) and Reasoning about Broadcasts K. V. S. Prasad Department of Computing Science Chalmers University of Technology S 412 96 Gothenburg Sweden E mail: prasad cs.chalmers.se. June 21, 1994 1 Background: First order CBS [Pra91, Pra93b, Pra93c] develop CBS, a CCS like calculus where processes communicate by broadcasting values along a single channel. Speech is autonomous, but hearing takes place only when the environment speaks. Each process has a unique response to each input it might receive. Non determinism arises only if processes ....

K. V. S. Prasad. A calculus of value broadcasts. In PARLE'93, June 1993. Springer Verlag LNCS 694.

....for PCBS has been implemented, and all the examples here have been run on it. Readers are assumed to be familiar with CCS [Mil89] and with functional programming. Familiarity with [Pra93c] is not strictly necessary, but would help. More discussion of the CBS model of computation can be found in [Pra93b, Pra91], which present older versions of CBS, significantly different from that of [Pra93c] 1. The Syntax and Semantics of PCBS A PCBS process of type P roc ff says or hears values of type ff and evolves to other processes of type P roc ff. Saying and hearing v: ff are written v k 00 and v k 000 , ....

....theorem below does not need to take care of priorities. Law 1(c) applied to the r.h.s. shows that all is well. Proposition 12 (Expansion theorem) Let r 2 f0; 1g. Then P 0 j P 1 P r;w;k w k (P 0 r j P 10r =w) x: P 0 =x j P 1 =x) where P r w k 000 P 0 r . CBS , the calculus of [Pra93b], is an unprioritised calculus with and prefixes, 0 and as primitive constructors. The equations 1(a) and 1(b) of Proposition 11 together with the expansion theorem constitute a complete axiom system for finite processes of CBS . This result has not yet been adapted to the CBS of ....

[Article contains additional citation context not shown here]

K. V. S. Prasad. A calculus of value broadcasts. In PARLE'93, June 1993. Springer Verlag LNCS 694.

No context found.

K. V. S. Prasad. A calculus of value broadcasts. In Parallel Architectures and Languages Europe, pages 391--402, 1993.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC