Question

Locked

what would be a good algorithm for improving VPN speeds

By atusiime ·
I want to develope a light algorithm that would enhance or improve Wide Area Networks VPN and them implement the out comes. How do I go about it?

I also want accademic papers for review in on the same.

Thnak you.
Annet.

This conversation is currently closed to new comments.

5 total posts (Page 1 of 1)  
| Thread display: Collapse - | Expand +

All Answers

Collapse -

Here you go

by IC-IT In reply to what would be a good algo ...
Collapse -

Message has been deleted.

by aoomall In reply to what would be a good algo ...
Collapse -

Free Best VPN Service Reviews

by bestvpn In reply to what would be a good algo ...

There are lot of service provider you can chose top service to get full speed vpn.

http://bestvpnservice.com

Collapse -

That's a rather ambitious goal

by robo_dev In reply to what would be a good algo ...

I would start by first reading and understanding the RFCs that define IPSEC, such as RFC 2401.

http://www.ietf.org/rfc/rfc2401.txt

Your premise is flawed.

The speed of the encryption algorithm is not the determining factor in VPN speed. Most VPNs use something like SHA-256 for hashing and AES for encryption. Are you planning to create your own hashing and encryption algorithms to replace SHA and AES?

Modern hardware-based encryption/decryption hardware are able to do real-time IPSEC encrpytion/decryption and hashing functions waaaay faster than the typical data rate over any public network. A Cisco ASA5580 can do 1GBps of throughput doing AES.

If the goal is to maximize the encrpytion/decryption rate, you throw more processing power at the problem, not reinvent the algorithms.

And obviously, in order to have an algorithm that's sufficiently secure, it has to require a certain amount of processing power to decrypt, or it would not be very secure.

While some modest VPN performance gains might be achievable with some more elegant data compression techniques, that's typically just not possible. For all intents and purposes, you cannot do 'post-process' compression of encrpyted data to get any real benefits; the two operations must happen at the same time.

Back to Networks Forum
5 total posts (Page 1 of 1)  

Related Discussions

Related Forums