You are currently browsing the category archive for the ‘5G’ category.

Yesterday Google Scholar sent me another alert about a new paper. I must say that Google Scholar is becoming my number 1 source to stay up to date about research in mobile security.

The paper, “Formal analysis of 5G authentication“, is a pre-print released by  a team from ETH Zurich, University of Lorraine and University of Dundee. Similarly to a recent paper on LTE security (LTEInspector: A Systematic Approach for Adversarial Testing of 4G LTE), the authors translate the 3GPP protocol specifications into pseudo-code that can be formally verified and analyzed. In this case, the authors analyze the recently released 5G 3GPP specifications, with special focus on the authentication protocols. To do so, the authors use Tamarin, a protocol verification tool.

I strongly recommend reading the paper. As I expected, the authors found a few weaknesses on the protocol. The 5G AKA protocol appears to fail to meet several security goals that are explicitly required by the 3GPP specifications, as well as other critical security properties. The paper highlights weaknesses in the standard and suggests improvements and refinements. Such an interesting work and an excellent paper.

It is worth noting that a couple months ago I was invited to write an opinion article on 5G security and I got some criticism from 3GPP folks on it, claiming that 5G is secure and things have been improved very much. As I stated in my article (Are we there yet? The long path to securing 5G mobile communication networks), I still see a long way to go to fully secure mobile communication networks. And the new sophisticated security architecture and PKI infrastructure are very interesting, but based on the unrealistic assumption that each SIM will have a public key or certificate for all operators from all countries. I always acknowledge that it is very hard to achieve a secure mobile communications system and the only reason I work in proactively identifying security weaknesses is to keep raising awareness on this problem.

It makes me happy to see so much excellent work coming from academia in the area of mobile security. Excellent research topic for talented PhD students to work on. And it makes me even happier that, just a couple of months after being publicly released, there is security research analyzing the 5G specifications. I am myself currently involved in a research project on 5G security with a team from VATech under Prof. Jeffrey Reed and Prof. Vuk Marojevic. We are working on a new paper on 5G security that should be out sometime later this summer or early Fall. Stay tuned! For the ones of you who saw me speak at UC Irvine last May or at Hushcon East in NY in June, you already got a bit of a sneak peak.


(Originally posted as an article on LinkedIn)

The mobile and wireless communication industry is highly susceptible, as are most sectors in the information technology industry, to drowning in a sea of buzzwords. “5G” is a concept that has been thrown around frequently for the past 6 years or so to define a futuristic – and potentially hard to achieve – connectivity scenario in which speeds of 1Gbps are ubiquitous, sub-10ms latencies are the norm, and the network can take on 1,000 times more connected devices without any hiccups. This utopian connected world has always been promised to arrive in 2020, to coincide with the Tokyo Summer Olympics, with the first trials during the 2018 Winter Olympics.

While the buzz around 5G has spawned conferences, workshops, symposiums, industry consortiums, and tomes of scientific press, some great minds in both academia and industry have been working on actual technology which, unlike big stands at expos and conferences and flashy slide decks, will solve the 5G connectivity challenges. mmWave communications are the clear path towards being able to achieve gigabit rates ubiquitously in dense urban scenarios and, although radio signal propagation is very challenging at such high bands, massive MIMO (Multiple-Input Multiple-Output) and adaptive beamforming arrays are the promising technologies that will help close that gap.

While 5G has mostly been a buzzword attached to flashy presentations and keynotes during the last few years, this does not change the fact that there have been outstanding research and development advances in some of the key technology areas that will sustain the connectivity demands of the next decade. That is, things that will make the concept of 5G an actual reality. As a result of this excellent work, the first official release of the 3GPP standards for 5G communication systems was published in December 2017. The new proposed mobile communication system is known as New Radio (NR) and its Core Network (as opposed to the Radio Access Network) is known as 5G System (5G-S).

While the technology pillars for future 5G mobile systems were being developed, there has been a spike in excellent security research work in the general field of mobile communications, and LTE mobile networks more specifically. As I anticipated 2 years ago, open source platforms have provided the perfect tools for bright security researchers to work on outstanding research projects that have yielded the discovery of all sorts of implementation issues and communication protocol deficiencies in LTE mobile networks. In some cases, the technology press has picked up on some of the resulting scientific publications at top conferences, which has sent shockwaves throughout the mobile communications industry. Such great research has also driven security innovation and protocol improvements that are making mobile networks nothing but more secure and resilient.

For quite a few years, I have been among the advocates for piggybacking on the technology disruption of 5G to address the well-known and, in many cases, very concerning security and scalability issues in LTE mobile networks. Although the major breakthrough in 5G will be at the physical layer (PHY), we are long overdue on reconsidering the current circuit-switched architecture of core mobile networks and embracing a fully packet-switched architecture. Although the mobile core of LTE is already fully IP-based, the architecture of the network still heavily relies on circuits – known as bearers in 3GPP jargon – and complex state machines. Among many other reasons for embracing a packet-switched architecture, the goal of massive connectivity in 5G networks will never be achieved in current control plane signaling-constrained networks. This is especially true when the goal is achieving connectivity for 1,000 times more devices and the Internet of Things (IoT) is at our doorstep, waiting to enter the game. As a great point of reference for this massive challenge in mobile networks, I always like to refer my colleagues to the visionary paper by J Kim and Paul Henry.

In general, the disruption of 5G is indeed the perfect opportunity for major architectural changes in the core network, though this is a challenging goal. However, it would be a big loss if, at the very least, 5G was not used to address the minor, and narrower in scope, changes required to tackle concerning security exploits uncovered in LTE. By now, it is well understood that there are multiple ways an adversary could abuse the pre-authentication Radio Resource Control (RRC) and Non-Access Stratum (NAS) messages, both of which are neither authenticated nor encrypted. As such, LTE mobile networks and, more importantly, LTE smartphones and network equipment, are potentially vulnerable to certain privacy leaks and Denial of Service (DoS) attacks, as prototyped in the lab by several research projects over the last 5 years.

The first release of the NR and 5G-S standards (Release 15 of the 3GPP standards), with the initial specifications released in December 2017, makes a partial attempt at addressing such security issues. Interestingly, most of the security definitions have not been included in the specifications until the updated documents released in March 2018. There are some ongoing efforts in protecting the International Mobile Subscriber Identifier (IMSI) using Public Key Infrastructure (PKI), likely motivated due to the recent amount of press and media coverage on IMSI catchers, in addition to leveraging PKI to authenticate certain pre-authentication messages. However, it is still to be seen how certain challenges, such as how to authenticate or implement PKI with a subscriber roaming from another network – or even a foreign network – will be solved. There are also several edge cases in which null integrity and null ciphering are used, such as the initial registration procedure for emergency services (3GPP TS 24.501 V1.0.0 2018-03 – Plus, the fact that null ciphering and null integrity are supported (3GPP TS 24.501 V1.0.0 2018-03 – Table could potentially end up in insecure, unexpected protocol edge cases. Besides that, the sheer number of pre-authentication messages still exposes protocols to potential security exploits.

I recently collaborated with a highly-renowned mobile security research team from academia (Prof. Jeffrey H. Reed and Dr. Vuk Marojevic at Wireless @ Virginia Tech) in a security analysis of the NR standards. In the past, both that team and I had been involved in research on protocol-aware jamming and the underlying vulnerability of LTE mobile networks to adversarial RF jamming. The goal of this latest security analysis was to investigate the feasibility of protocol-aware jamming in the proposed PHY layer in NR. The outcome of the study will be presented in the 1st IEEE Workshop on 5G Wireless Security coming up this May in Kansas City, but the results are already available to the public in our paper.

Although the outcome of the security analysis is not encouraging, one must acknowledge that it would have been a massive achievement to simultaneously tackle the challenge of gigabit connectivity, mmWave combined with massive MIMO and, on top of that, security and resiliency. Things at the higher protocol layers still look rather challenging as well. Despite my forays into PHY layer security and protocol-aware jamming, most of my security research work over the last 8 years has focused on protocol-level exploits on various wireless technologies, with great focus on 3GPP’s LTE as defined starting on Release 8. As one of the few researchers who has uncovered numerous protocol exploits that would result in DoS of mobile devices and privacy leaks, I was, and still am, optimistic about the role of 5G disruption in enhancing the security of mobile networks. And leveraging PKI is definitely a step in the right direction.

Nevertheless, what we have seen so far in the NR and 5G-S standards is not a complete solution yet, despite attempts to address protocol exploits. In parallel, there is also a set of new pre-authentication messages and new fields and configurations in existing messages. One must acknowledge, though, that fully securing mobile communication networks is a massive challenge that will require collaboration among academia, industry and researchers.

As a security researcher, many of my colleagues and I see the emerging landscape of 5G as a blank canvas to experiment with the potential security impact of adversarial tampering, spoofing and intercepting of these pre-authentication messages in NR and 5G-S. And it is critical that the security research community and the mobile communications industry work together in identifying such potential exploits and, more importantly, their root causes, so the security of the upcoming 5G networks can be enhanced in short order. There is still time until 2020 to enhance mobile communication networks even further.

Although we could – and perhaps should – be much closer by now, there is still a very long way to go to fully secure mobile communication networks. The same applies to reaching a flexible and truly scalable mobile architecture capable of supporting the connectivity demands of the future. However, the very active community of mobile security researchers will hopefully take us to that stage.

Equipped with an army of software radios, mostly USRP B210s, and my new toolset based on srsLTE, I continue my work on protocol-fuzzing mobile and wireless network standards with the goal of contributing to the security of the communication systems used by billions on a regular basis. In my case, time is now slightly scarcer due to my recent fatherhood. But, I continue to work and follow closely the excellent work of the few other teams in this research field, where outstanding graduate students and researchers are paving the way towards secure and reliable mobile communication networks.

This time, I cannot predict whether there will be a large number of new security exploits identified and prototyped in NR and 5G-S networks in the near future or a spike in mobile security findings in this field, mainly because there are no available test-beds or open source implementations of the Release 15 stack. But, as long as the folks behind tools such as srsLTE keep up their great work, it will not take long for the right tools to be available for applied security research on 5G mobile systems. And when that day comes, it is game on!

Roger Piqueras Jover is a Security Architect in the Office of the CTO at Bloomberg, where he is actively engaged in mobile and wireless security research. He maintains a bibliography of his previously released and published work at his personal website:

About me:

Born in Barcelona, moved to Los Angeles at age 24, ended in NYC, where I enjoy life, tweet about music and work as a geek in security for wireless networks.
All the opinions expressed in this blog are my own and are not related to my employer.
About me:

Blog Stats

  • 133,353 hits

Twitter feed

Enter your email address to follow this blog and receive notifications of new posts by email.