HELP:Cannot perform the operation: {real affine} .* {convex}

Nonconvex
1 
function [p,r,Vk] = MaxSumRate_CVX(h,No,P_max,K,J,Nt,Nr)
% Maximizing system sum rate using CVX
    cvx_begin 
    cvx_solver mosek
        variable p(K,J) nonnegative % The power of each sub message
        variable Vk(Nt,J) complex % The digital precoding metrix
        dual variables w u
        expressions r(K,J) h_gain(K,J) SINR(K,J)
        h_gain = []; % Channel gain
        for k=1:K
            gain{k} = sum(pow_abs(h{k}*Vk,2));
            h_gain = [h_gain; gain{k}];
        end
        SINR = (p.*h_gain)./No; % Calculate SINR
        for k=1:K
            for j=1:J
                r(k,j) = log(SINR(k,j) + 1) ./ log(2);
            end
        end
        maximize sum(r(:))
        subject to
            w:sum(p(:)) <= P_max;
            w:sum(p,1) <= P_max/K;
            u:square_pos(norm(Fk)) <= 1;
    cvx_end
end

Disciplined convex programming error:
Cannot perform the operation: {real affine} .* {convex}

出错 MaxSumRate_CVX (第 14 行)
SINR = (p.*h_gain)./No; % Calculate SINR
I don’t know how to solve it.

2

p is a real affine expression (scalar) , and h_gain is a cvx convex expression (scalar), h_gain is just a scalar metrix, how can I transfer h_gain to be a real affine expression (scalar)?

3

You are multiplying the convex expression h_gain by another CVX variable, p. That is a non-convex operation.

Have you proven this is a convex optimization problem with this choice of optimization variables?