A007224 Number of distinct perforation patterns for deriving (v,b) = (n+3,n) punctured convolutional codes from (2,1).
2, 9, 38, 143, 546, 2066, 7752, 29070, 108968, 408595, 1533870, 5766243, 21710850, 81880920, 309328008, 1170524970, 4436618940, 16842720336, 64037794548, 243836217702, 929759970392, 3549992610584, 13571935767600, 51950354135888
Offset: 4
Keywords
References
- Guy Bégin, On the enumeration of perforation patterns for punctured convolutional codes, Séries Formelles et Combinatoire Algébrique, 4th colloquium, 15-19 Juin 1992, Montréal, Université du Québec à Montréal, pp. 1-10.
- N. J. A. Sloane and Simon Plouffe, The Encyclopedia of Integer Sequences, Academic Press, 1995 (includes this sequence).
Links
- Guy Bégin and David Haccoun, High rate punctured convolutions codes: Structure properties and construction techniques, IEEE Transactions on Communications 37(12) (1989), 1381-1385.
- David Haccoun and Guy Bégin, High rate punctured convolutional codes for Viterbi and sequential coding, IEEE Transactions on Communications, 37(11) (1989), 1113-1125; see Section II.
Programs
-
Maple
with(numtheory):P:=proc(b,v0) local k: RETURN(add(phi(k)*(1+z^k)^(v0*(b/k)),k=divisors(b))/b): end; seq(coeff(P(b,2),z,b+3),b=4..40); (Pab Ter)
-
Mathematica
a[n_] := DivisorSum[GCD[n+3, n], EulerPhi[#] Binomial[2n/#, (n+3)/#]&]/n; Table[a[n], {n, 4, 40}] (* Jean-François Alcover, Mar 09 2023 *)
Formula
From Petros Hadjicostas, Jul 27 2020: (Start)
The number of perforation patterns to derive high-rate convolutional code (v,b) (written as R = b/v) from a given low-rate convolutional code (v0, 1) (written as R = 1/v0) is (1/b)*Sum_{k|gcd(v,b)} phi(k)*binomial(v0*b/k, v/k).
According to Pab Ter's Maple code, this is the coefficient of z^v in the polynomial (1/b)*Sum_{k|b} phi(k)*(1 + z^k)^(v0*b/k).
Here (v,b) = (n+3,n) and (v0,1) = (2,1), so
a(n) = (1/n)*Sum_{k|gcd(n+3,n)} phi(k)*binomial(2*n/k, (n+3)/k).
This simplifies to
a(n) = (1/n)*(binomial(2*n, n+3) + 2*[(n mod 3) == 0]*binomial(2*n/3, (n/3) + 1)). (End)
Extensions
More terms from Pab Ter (pabrlos2(AT)yahoo.com), Nov 13 2005
Comments