Here's another way, basically a variant of what roboticus posted (the key point being to use an array to keep the target words in one easily maintained place). This one uses a single capturing regex to handle all terms at once, and reports the hits instead of the misses:

use strict;
use warnings;

my @required = qw( name timestamp dl_retransmits_pct dl_throughput mac
+ );
my $req_regx = join( '|', @required );

while (my $line = <DATA>) {
    my %matched = map{ $_ => undef } $line =~ /($req_regx)/g;
    print $line if scalar( keys %matched ) == scalar( @required );
}

__DATA__
mac timestamp name dl_retransmits_pct dl_throughput is a hit
mac timestamp name dl_retransmits_pct throughout is a miss
name timestamp dl_retransmits_pct dl_throughput is a miss
foo mac name dl_throughput name timestamp name dl_retransmits_pct mac 
+mac is a hit
name timestamp dl_retransmits_pct is a miss
dl_throughput is a miss
[download]

But as Fletch suggested, if your input is really JSON data, you should use a JSON parser and handle the input as the data structure that it really is.

(updated to fix grammar and spelling)