| E. Rains, private communication, December 1996. 15 FIGURES 0.095 0.167 0.25 0.5 |
....and Rains [5] extend to arbitrary ; and allow for small deviations of the boundary densities as in (5.2) with n replaced by v . Very recently they investigate the case of re ection symmetry relative to the diagonal, in particular = and an extra line density along the diagonal [15]. The particle model behind the Poisson last passage percolation is the PNG model [16, 17, 18] It consists of point particles with velocity 1. They annihilate each other at a collision and are created in pairs with rate 1. If (x; t) is the density of right movers, and (x; t) of left ....
....at the lower edge the w(j; 0) are exponential with rate 1 . By maximizing the passage time on scale n we obtain the same phase diagram as for the critical step. However the uctuations in N t , now the number of particles injected up to time t, are modi ed. In Figure 4 we summarize the ndings [15] which could be written more formally as in Conjecture 7.1. GSE is the distribution of the largest eigenvalue of a symplectic Gaussian random matrix. F 0 is a novel distribution and given by 0 (x) 1 ( v(x) u(x) x 2u (x) 2u(x) F (x)E(x) 7.15) in the notation of [5] So ....
J. Baik, E.M. Rains, private communication
....is zero if ffi 1=6 [20] It remains an open question whether the 1 Gamma H(ffi) Gamma ffi log 2 3 bound also applies to degenerate codes. Very recently, it has been announced that the ffi 1=6 threshold bound does extend to nondegenerate codes, and this will be appear in a forthcoming paper [22]. It is interesting to note that there exist some degenerate stabilizer codes that outperform all known nondegenerate codes on the depolarizing channel, for some values of ffi [5] The best lower bound for this channel that we are aware of is 1 Gamma H(2ffi) Gamma 2ffi log 2 3 [8] For the ....
E. Rains, private communication, December 1996. 15
....which protects 1 qubit of information against 2 errors using n = 9 qubits. A solution of the Macwilliams identities exists for codes mapping 1 qubit into n = 10 qubits; however an extension of the techniques in this paper based on classical shadow code techniques rules this possibility out as well [20]. The smallest possible code protecting against two errors thus would map 1 qubit into n = 11 qubits; such a code was constructed in [15] Both possibilities eliminated above would have required degenerate quantum codes. These might have allowed to find more compact codes than would have been ....
....such a code was constructed in [15] Both possibilities eliminated above would have required degenerate quantum codes. These might have allowed to find more compact codes than would have been expected from an analogy to classical codes. A systematic study of the MacWilliams identities for n 30 [20] shows that this is not the case. The most compact codes appear not to be degenerate. It will be interesting to know if this holds as n 1. In conclusion, we have derived the quantum analog of the MacWilliams identities which give necessary conditions for the existence of codes. We have ....
E. Rains, private communication.
No context found.
E. Rains, private communication, December 1996. 15 FIGURES 0.095 0.167 0.25 0.5
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