You might want to look into socket programming in Perl; it's not really that terribly difficult to get started, and there are some good examples around the Monastary. Here is a short example of networking with IO::Socket.

use strict;
use IO::Socket;

my $host = 'pdp11.mydomain.net';

my $port = 6740;

my $socket = new IO::Socket::INET(PeerAddr => $host,
                                  PeerPort => $port,
                                  Proto    => 'tcp');
die "cannot open socket" unless ($socket);

#send a custom command to the PDP-11 server

my $cmd = "GIMME\n";
print $socket $cmd;

# print the server's response to the gimme request

while (<$socket>) {
    print;
}
close $socket;
[download]

The reason I suggest this type of approach is that using netcat is going to limit portability for your code, if you decide to try and run it on a platform that doesn't have netcat for instance.

Also, using spawned external child processes will likely lead to more issues such as the one you're having. It's probably a better idea to use the functionality that's already built in to Perl.