Chinese state media outlet, Xinhua, reported
yesterday that the world's first mimic DNS (Domain Name System) server went
into operation Tuesday in Central China's Henan province. This was described as
a proactive move to prevent cyberattacks.
According to Xinhua, the Henan Communication Administration revealed
that the server is in use by China Unicom's Henan provincial branch. It marks
the first practical application of the mimic security theory proposed by
Chinese scientists.
The mimic DNS server will not change the current network
structure but can effectively prevent various known and unknown attacks
targeting the system.
Wu Jiangxing, a Chinese Academy of Engineering academician
and head of the research team, said, “Mimic DNS server is just our first
application. The mimic web server, mimic cloud, mimic data center and other
network devices will be launched in the future.”
The technology is said to be inspired by Mimic Octopus, which
is an expert in camouflage, changing appearance to adapt to the environment.
Chinese scientists proposed the idea of mimic computing in
2007 and produced
the world's first mimic computer prototype in 2013.
What is mimic defence
technology?
A 2016 paper (pdf
download) from researchers at the PLA Information Engineering
University, a university-level military academy, explained the concept of mimic
defence technology. It talks about different types of active defence technologies,
which seek to prevent attacks or minimise their impact, as opposed to traditional
defence methods, such as intrusion detection, virus detection, and firewalls, which
are usually about responding to attacks.
According to the paper, the early form of active defence
technology is intrusion tolerant technology. An intrusion tolerant system can
continue to work correctly and provide the expected service to users, despite
successful attack on some components. However, there is limited research
regarding intrusion tolerance, and redundancy and voting may lead to high
resource costs and time delays.
Then there is moving target defence which seeks to use time-varying
migration and changing mechanisms and strategies across multiple system
dimensions, to limit exposure of vulnerabilities, reduce attack opportunities,
and increase the cost of attack significantly. The randomness, and unpredictability
created by changing the configuration, composition, or state of the system
dynamically, makes it difficult for attacked to conduct effective attacks. However,
for this to be effective, it is necessary to gain a high changing frequency,
which may cause some loss in the system performance.
The paper says that mimic defence technology “introduces a hybrid scheduling strategy
using dynamic heterogeneous redundancy construction, and uses the
re-constructional, reconfigurable, re-definable, and virtualised construction
methods of dynamic heterogeneous redundancy to enhance the uncertainty so that
the difficulty in detection increases nonlinearly, and the attack is
transformed into a minimal probability event.”
Mimic defence is supposed to provide immunity against unknown
vulnerabilities, traps or backdoors. It can be integrated with passive defence
means to protect against known or unknown attacks in cyberspace. The paper goes
on to caution that mimic defence is not attempting to solve all cyberspace
security problems and it does not exclude the integration of proven defence
measures.
The Ministry of Science and Technology of the People’s
Republic of China commissioned a study for the testing mimic defence principle
verification systems. The tests were organised by the Science and Technology
Commission of the Shanghai Municipality at nine testing facilities from January
to June 2016, including the National Research Center for Information Technology
Security. Applications in two areas were considered, a mimic router and a mimic
web server.
In order to test the endogenous defence mechanism of the
mimic defence system, the tested objects could not install any protection
tools or carry out vulnerability repair or backdoor blocking. The object also cannot
use protections such as firewalls, encryption authentication, etc. The testing
used a variety of methods and means to answer questions such as whether the system
can conceal the unknown vulnerabilities and backdoors, whether attackers can
exploit unknown vulnerabilities to inject unknown Trojan virus and whether the
running environment allows the “toxic carrier” in the mimic system.
The outcome was viewed as a successful representation of the
application of the mimic defence theory.