6.7 An Application to Cyclic and BCH Codes

1.
a. If n = 1 it is clear, if n = 2, img. In general, img
c. If f = a0+ a1x + img then f′ = a1 + a3x2 + a5x4 + img = g(x2) where g = a1+ a3x + img. Conversely, f = g(x2) = b0+ b1x2 + b2x4 + img clearly implies that f′ = 0 (as 2 = 0 in img
3. If n = 2k then

img

and, by induction, img; that is 1 − xn = (1 + x)n. So the divisors of 1 − xn are 1, (1 + x), (1 + x)2, . . ., (1 + x)n, and these are a chain under divisibility.
Conversely, if 1− xn = (1 + x)kpm img where p is irreducible and p ≠ (1 + x), then neither 1+ t nor img contains the other in Bn. If n = 2km, m odd, then img a contradiction as m = 1 in img
5.
a In B4 : 1 + t, t + t2 = t(1 + t), t2 + t3 = t2(1 + t) and 1 + t3 = t3(1 + t). The other members 0, 1 + t2, t + t2 and 1 + t + t2 + t3 all lie in smaller ideals. (See Example 3.)
6.
a. We have |D| = 2 so CD = 0 or D. If n is odd then 1 + t + img + tn−1 has odd parity, so CD = 0 in this case. Also, if n − 1 = 2m then

img

is in C. Thus tn−1 img (C + D) − C. Since |C| = 2n−1 implies C is maximal, C + D = Bn. Thus BnC × D when n is odd by Theorem 7 §3.4.
7.
a. 1 + x7 = (1 + x)(1 + x + x3)(1 + x2 + x3) so there are 2 · 2 · 2 = 8 divisors in all. Thus there are 7 codes (excluding img).
c. 1 + x8 = (1 + x)8 so there are 9 − 1 = 8 codes in B8.
e. 1 + x10 = (1 + x2)(1 + x2 + x4 + x6 + x8) = (1 + x)2(1 + x + x2 + x3 + x4)2 (using Exercise 10 below). Hence there are 3 · 3 − 1 = 8 codes in B10.
9. Since ui img E and |E| = n, we have (ui)n = 1 by Lagrange's theorem. Then ui is a root of xn − 1, whence mi divides xn − 1.
10.
a. Since g = 1 + x + x2 + x3 + x4 has no root in img, if it factors, it does so as

img

Hence aa′ = 1 = cc′ so a = a′ = 1 = c = c′. Hence

img

so 1 + bb′ + 1 = 1 (coefficient of x2) whence b = b′ = 1. But then the coefficient of x is 1 = b + b′ = 0, a contradiction.
11. Since g = 1 + x + x4 has no root in img, if it factorizes, it does so as

img

Thus aa′ = 1 = cc′ so a = a′ = 1 = c = c′. Thus g = (1 + bx + x2)(1 + bx + x2) so (coefficient of x3) b + b′ = 0 and (coefficient of x) b + b′ = 1. This is impossible.
13. (1 + x + x4)(1 + x3 + x4)(1 + x + x2 + x3 + x4)

    = (1 + x + x3 + x4 + x5 + x7 + x8)(1 + x + x2 + x3 + x4)

    = (1 + x3 + x6 + x9 + x12) .

Since (1 + x)(1 + x + x2) = 1 + x3 = 1 − x3, this verifies the factorization. We have checked that 1 + x + x4 and 1 + x + x2 + x3 + x4 are irreducible. If 1 + x3 + x4 = (a + bx + cx2)(a′ + bx′ + cx2), then aa′ = 1 = cc′ so a = a′ = 1 = c = c′. Thus 1 + x3 + x4 = (1 + ax + x2)(1 + ax + x2) so (coefficients of x, x3) a + a′ = 0 and a + a′ = 1, a contradiction.
15. Write m = 1 + x + img + xn−1. Let img where g divides (xn − 1). If f(t) img C has odd parity, then f(t) = q(t)g(t) so 1 = f(1) = q(1)g(1). Thus g(t) has odd parity. We must show that m(t) img C. Let gh = xn − 1 so

img

Since x + 1 is prime, either (x + 1) g or (x + 1) h. But g = λ(x + 1) gives g(1) = λ(1)(1 + 1) = 0, a contradiction. So h = k(x + 1), whence (*) gives gk(x + 1) = (x + 1)m; m = hg; m(t) img g(t).
17.
a. By Theorem 10 §4.2, let 1 = qg + ph in img and define e = qg. Then e(t) img C so e(t) ⊆ C. On the other hand 1 = e + qg so g = eg + p(1 − xn). This means g(t) = e(t)g(t) img e(t), so C = e(t). Since e(t) = q(t)g(t), this gives e(t)2 = q(t)e(t)g(t) = q(t)g(t) = e(t).
c. We have

img

We have 1 + x + x2 + x4 = x(1 + x + x3) + 1 so

img

So take e = x(1 + x + x3) = x + x2 + x4. Note that

img

So e(t) = t + t2 + t4 is the idempotent generator.
e. Given e(t) in Bn, e(t)2 = e(t2); img. If n = 2k, and e(t)2 = e(t), this gives e(t) = e(t)n = e(tn) = e(1) = 0 or 1. In B6, e(t) = 1 + t2 + t4 is idempotent because

img

18.
a. Given f(t) = a0 + a1t + img + an−1tn−1 in Bn, write img for the corresponding word. Then img so img if and only if f(ui) = 0 for all i. This is certainly true if img the converse holds because the roots are distinct.
19. The roots of G are linearly independent over img by choice, so G has rank k. Since rank R = rank G, it follows that R = [InA] by the definition of a row echelon matrix. To see that R generates C, it suffices to show that the rows of R span C. Since the rows of G span C, it suffices to show that the rows of G are all linear combinations (over img of the rows of R. But R = UG, U invertible, so G = U−1R and the result follows.
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
18.118.9.7