8 Effective Algorithms to Wipe and Erase Data Permanently

In contrary to protecting data, people in many cases need to wipe their sensitive or useless data. And many data destruction programs appear and use different algorithms to help with this task. If you wonder how these algorithms work, you can have a look at this article.

If you want to erase some sensitive data or remove all the data before selling your used storage device, then wiping data thoroughly is of great importance. Criminals often do everything to steal your personal information for illegal profit. And halfway data scrubbing will be the vulnerability that offenders can make use of. Thus, For example, bad persons can use software to attempt a pst recovery, then they can find all the data in your Outlook.

Since people are likely to use third party tools to help them with the task, today we will introduce 8 common algorithms  in qualified tools to help you choose an up-to-standard one.


1.DoD 5220.22-M

This method is the most common algorithm used by plenty of data destruction tools. And it is defined by the US National Industrial Security Program (NISP) and works in following way.

  • Write a 0 and verify
  • Write a 1 and then verify
  • Write a random character and then verify


This method is originally defined in the Forest Green Book. And it is published by a group which is ever part of the US National Security Agency (NSA). And it works exactly the same as the DoD 5220.22-M method, but offers people more choice like increasing more overwrite processes.


This method is defined by the United States Air Force (USAF) and also uses 1, 0, random character. But it only implement with verifications at the last pass which is different with above methods.

  • Write a zero
  • Write a one
  • Write a random character and then verify

4.AR 380-19

This method is published by the US Army and defined in Army Regulation 380-19. And it has a totally different ways from foresaid means.

  • Write a random character
  • Write a typical character
  • Write the specified character complement and then verify

If you find it applied without a final pass verification or a third pass, then the tool is unqualified.

5.NAVSO P-5239-26

This method is published by the US Navy. In many data destruction programs, it works in following ways.

  • Write a specified character
  • Write the typical character complement
  • Write a random character and then verify

But on the basis of the actual specification, this work pattern proves less effective than the other complicated overwrite patterns.


This method is originally published by the Royal Canadian Mounted Police (RCMP). And it applies more overwrite processes in following ways.

  • Write a 0
  • Write a 1
  • Write a 0
  • Write a 1
  • Write a 0
  • Write a 1
  • Write a random character and then verify

More overwrite processes increase the success of data wiping. But at the same time it will cost you a lot of time to finish the task. So you can select as per you like.


This method is published by the Communications Electronics Security Group (CESG). And it combines 2 kinds of overwrite patterns called “HMG IS5 Baseline” and “HMG IS5 Enhanced”. And they work in following ways.

HMG IS5 Baseline

  • Write a 0
  • Write a random character and then verify

HMG IS5 Enhanced

  • Write a zero
  • Write a one
  • Write a random character and then verify

8.ISM 6.2.92

You can find the latest version of the ISM in the website of “Australian Government Department of Defence”. And of course this method is originally issued by the Australian Department of Defense: Intelligence & Security.

This method works in following ways.

  • Write a random character and then verify

Besides, this method specifies that if the capacity of a drive is under 15 GB, then the drive needs at least 3 times overwrite to wipe the data permanently.

Author Introduction:

Allissa Shen is a data recovery expert in DataNumen, Inc., which is the world leader in data recovery technologies, including fix SQL Server and word recovery software products. For more information visit www.datanumen.com

Comments are closed.