The following post is based from this article here.
In this article from Ars Technica, discusses a few topics that I have been recently thinking about. The first is to do with the security of the quantum information at its endpoint, and the second is to do with the scope of distance that in which QKD can be utilized.
Whilst QKD is known for its security against interception and copying due to its quantum physics properties, I have wondered about how security is ensured at its destination. Now, while this may seem redundant; why have security for place the key is supposed to reach? I automatically thought of this due to the strength of QKD.
The author of the article, Chris Lee, mentions end point security in relation to long distance QKD. This is because, QKD currently has a range of approximately 70-100km, and hence, a long distance QKD would require intermediaries. While the two final endpoints may be secure, this isn’t necessarily the case for the intermediate hosts.
The article states that research has been conducted with the intent of resolving this weakness. To clarify, the main weakness here is that current quantum computers are ‘transparent’, to quote the article, in how they store their information. So, while the traversing data is secure, the data stored in the quantum computers at the endpoint can be easily accessed. ‘Easy’ being a relative word here. The solution to this, is to encrypt the quantum memory stored upon the computers.
This encryption is quite fascinating as it uses an intrinsic quantum principle known as Heisenberg’s Uncertainty principle (Link here for more information.) and frosted glass. The article describes this encryption method as follows. A control light has its intensity randomly defined throughout space by shining it through a piece of frosted glass. The quantum bit of qubit can also be defined as a pulse of light. However, in defining the qubit into a pulse of specific arrival time, the Energy, and hence, the wavelength of the pulse becomes uncertain, to the extent of being undefined. This can be shown mathematically as follows:
By employing the use of the control light dispersed through the frosted glass, the qubit pulse’s information can be stored within a cloud of spatial and light intensity dependent probability. The encrypted information then requires a light that matches both the light intensity and spatial pattern of the control light, for it to be decrypted. The article states that the accuracy of the decrypting light needs to be within 0.1% deviation of the encrypting control light to be effective.
The article continues on to further explain the issues involved in this sort of encryption, and the potential use of crystals.
I found this article to be very intriguing in that it looked into a flaw related to quantum mechanics and QKD, and suggested a potential solution. As the use of QKD increases, I consider that the need for preemptive determination of flaws is incredibly important.