T1(n)=O(f(n)) => T1(n) < c1 * f(n) where c1 is +ve constant (from Big O Def)
T2(n)=O(f(n)) => T2(n) < c2 * f(n) where c2 is +ve constant (from Big O Def)
a] True
Since T1(n) + T2(n) < (c1+c2) * f(n), So we can say T1(n) + T2(n) = O(f(n))
b] False
Let T1(n) = n2 and T2(n) = n and f(n) = n3
Clearly T1(n) >T2(n),So T1(n) = O(T2(n)) is not possible.
c] False
Let T1(n) = n and T2(n) = n2 and f(n) = n3
Clearly T1(n) < T2(n),So T1(n) = ώ(T2(n)) is not possible.
d] False
Take the same assumptions as above then
You can say, T1(n) < C1 * T2(n) for C1 = 1 but
T1(n) > C2 *T2(n) is not possible for any positive constant
Then T1(n)= Θ(T2(n)) is not possible.