Tsirelson showed that 2 √ 2 is the maximum value that CHSH expression can take for quantumcorrelations [B. S. Tsirelson Lett. Math. Phys. 4, (1980) 93]. This bound simply follows from the algebra of observables. Recently by exploiting the physical structure of quantum mechanics like unitarity and linearity, Buhrman and Massar [H. Buhrman and S. Massar Phys. Rev. A 72, (2005) 052103] have established that violation of Tsirelson's bound in quantum mechanics will imply signalling. We prove the same with the help of realistic joint measurement in quantum mechanics and a Bell's inequality which has been derived under the assumption of existence of joint measurement and no signalling condition.
IntroductionThere exists quantum-mechanical states shared between two parties which exihibit nonlocal character. This nonlocality is quantified by using 'Bell's expression'. This is an expression which is bounded by a certain value for 'Local Hidden Variable (LHV) models'; but can exceed this value in case of quantum correlations. Consider for example a setting of two parties, Alice and Bob; sharing a quantum state ρ and each has a choice of two local measurements. Alice can measure the observables A and A ′ whereas Bob's observables are B and B ′ . The measured values of all the obsevables can be 1 or -1. One relevant Bell's expression in this case is the Clauser-Horn-Shimony-Holt (CHSH) expression [2]. For local hidden varriable models, this expression is bounded by 2 but in case of entangled quantum systems,this bound can be violated. For example, on the singlet state of two qubits there exist observables (A,A ′ ,B,B ′ )for which value of the above expression is 2 √ 2. In fact as shown later by Tsirelson [3] that 2 √ 2 is the maximum quantum value of the 1 CHSH expression.Tsirelson's bound is a simple mathematical consequence of the axioms of quantum theory, but it would be interesting to know that whether there is some deeper reason why a violation greater than 2 √ 2 is unphysical. It is known in this connection that a violation greater than 32 3 ≃ 3.27 would imply that any communication complexity problem can be solved using a constant amount of communication [4].But this does not answer the question that what odd would have happened for a violation just greater than 2 √ 2. Recently by exploiting the physical structure of quantum mechanics like unitary dynamics and linearity; Buhrman and Massar [1] have shown that exceeding Tsirelson's bound by quantum mechanics will imply signalling in quantum mechanics. Here we provide a simple proof of the same by exploiting nice results of existence of joint measurement for spin along two different directions in quantum mechanics [5,6,7,8,9] and a Bell's inequality derived under assumptions different than that of the local-realism.
Joint measurement, No signalling and Bell's InequalityUsually, Bell's inequality is derived under the notion of local-realism. So, its violation by a theory will imply that the theory is incompatible with the notion of local-realism. For example, ...