I need to replace all the occurrences of  with 0x02 using sed. Will the following work or I'll have to use \?
sed -i -E "s/<>/<0x02>/g" $output_file
If yes, how can I do that?
I need to replace all the occurrences of  with 0x02 using sed. Will the following work or I'll have to use \?
sed -i -E "s/<>/<0x02>/g" $output_file
If yes, how can I do that?
There are no special characters in the string  that needs to be escaped for the string to be interpreted literally in the pattern part of a s/// command in sed.
There are furthermore no special characters in the replacement string 0x02 that needs to be treated specially (& would have been special in the replacement, as would \1, \2 etc. and \n, \t etc. if you're using GNU sed).
This means you should be able to just do
sed 's//0x02/g' inputfile >outputfile
The only thing that may need to be treated especially is to avoid replacing  when it occurs in substrings such as #. Since I don't see any example document in the question I can't say exactly if this would be an issue and how to best protect against it.
In general, don't use the -i option with sed until you have a sed editing expression or script that you know works, to avoid accidentally loosing data. You also don't need -E here as we're not dealing with extended regular expressions at all.
If you are running the sed command on a file whose name is kept in a variable, remember to double quote the expansion of that variable (see Why does my shell script choke on whitespace or other special characters?).