Here's one way that could work, if you have enough memory to generate the idx matrix, which is of size [size(A,1) x size(B,1)]
% crating some tables
rng(1)
Time = dateshift(datetime('now'),'start','day')+seconds(0:86399).';
StartTime = Time(1)+hours(0:23).';
n = numel(StartTime);
EndTime = StartTime+minutes(randi([1,120],n,1));
Value = rand(n,1);
A = table(Time);
B = table(StartTime,EndTime,Value);
A,B
% initialize A.Value
A{:,'Value'} = NaN;
% get the index of the row in B whose interval covers the entry in A.Time
idx = A.Time >= B.StartTime.' & A.Time <= B.EndTime.';
% r is row index in A, c is row index in B
[r,c] = find(idx);
In case intervals in B can overlap, if you want to take the first B.Value for A.Value do this:
A.Value(r(end:-1:1)) = B.Value(c(end:-1:1));
Or if you want to take the last B.Value in overlapping intervals for A.Value (or if you don't care, or if there are no overlaps - in which case it doesn't matter) do this:
A.Value(r) = B.Value(c);
For example, two intervals in this B overlap from 10:00 to 10:05
% taking the B.Value from the first overlapping interval (9:00 to 10:05)
% for A.Value from 10:00 to 10:05
A.Value(r(end:-1:1)) = B.Value(c(end:-1:1));
A(10*3600:10*3600+3,:)
A(10*3600+300:10*3600+300+3,:)
% taking the B.Value from the last overlapping interval (10:00 to 10:51)
% for A.Value from 10:00 to 10:05
A.Value(r) = B.Value(c);
A(10*3600:10*3600+3,:)
A(10*3600+300:10*3600+300+3,:)
