import sys
input = lambda : sys.stdin.readline().rstrip()

sys.setrecursionlimit(2*10**5+10)
write = lambda x: sys.stdout.write(x+"\n")
debug = lambda x: sys.stderr.write(x+"\n")
writef = lambda x: print("{:.12f}".format(x))


# FFT
import numpy as np
TYPE = np.int64
M = 998244353
def fft(a,b):
    l = len(a) + len(b) - 1
    l = 1<<((l-1).bit_length())
    c = np.fft.irfft((np.fft.rfft(a,l))*(np.fft.rfft(b,l)),l)
    c = np.rint(c).astype(TYPE)
    return c
def fft_large(a,b):
    d = 30000
    a1, a2 = np.divmod(a,d)
    b1, b2 = np.divmod(b,d)
    aa = fft(a1,b1) % M
    bb = fft(a2,b2) % M
    cc = (fft(a1+a2, b1+b2) - (aa+bb)) % M
    h = (((aa*d)%M)*d  + cc*d + bb) % M
    return h
def fft_large(a,b):
    """精度が足りないときはこちら
    """
    d = 1<<10
    a1, a2 = np.divmod(a,d*d)
    a2, a3 = np.divmod(a2,d)
    b1, b2 = np.divmod(b,d*d)
    b2, b3 = np.divmod(b2,d)
    aa = fft(a1,b1) % M
    bb = fft(a2,b2) % M
    cc = fft(a3,b3) % M
    dd = (fft(a1+a2, b1+b2) - (aa+bb)) % M
    ee = (fft(a2+a3, b2+b3) - (bb+cc)) % M
    ff = (fft(a1+a3, b1+b3) - (aa+cc)) % M
    h = (((aa*d*d)%M)*d*d + ((dd*d*d)%M)*d + (bb+ff)*d*d + ee*d + cc) % M
    return h

n = int(input())
a = list(map(int, input().split()))
b = list(map(int, input().split()))
a = np.array(a, dtype=np.int64)
b = np.array(b, dtype=np.int64)
aa = fft_large(np.arange(1,n+1).astype(np.int64), a)
bb = fft_large(np.arange(1,n+1), b)
ans = [0] + (aa+bb)[:2*n-1].tolist()
write(" ".join(map(str, ans)))