Another author of this presentation is Seungjoo Kim.
In this presentation, we prepared together, but only I speak.
The title is Blockchain for Cyber Defense and Will It Be Good As You Think? is our topic.
It's 5 a.m. in Korea, so I think the condition is not normal, but I hope you understand.
Again, I'm Seungri Lee, and I'm a Ph.D. student at Korea University.
I'm the captain of the Republic of Korean Army.
I have three years' experience as a researcher in the Agency for Defense Development.
It's about the development of the defense and armed system.
As a security researcher, I'm interested in network security and cyber defense.
About blockchain, I was a speaker of CryBlock 2019 at IEEE Infocom.
The title was Countering Block Withholding Attack Efficiently.
Also, I will speak at CryBlock 2020 at ACM Mobicom.
It will be September, and the title will be No Prosperous Take Has Taken Predatory Destructive Attack on Prosperous Take Cryptocurrencies.
It will be about a novel attack methodology about POS cryptocurrency.
It's not about research. I like to get groundbreaking ideas from philosophy.
My favorite philosophers are Derrida, Nietzsche, Foucault, and Wittgenstein.
Another author of this presentation is Seungju Kim.
He's my advisor, and he's a professor of cybersecurity at Korea University.
He's a reviewer in several international hacking conferences like Black Hat Asia and Secure Insight.
He's also an advisor of an undergraduate hacking club, Cyker.
The hacking group Cyker is now competing in DEF CON online.
His main research areas are focused on trustworthy systems, development methodology, secure SDLC, RMAF, common criteria, CMVP, blockchain, etc.
So our topics are, what is blockchain for cyber defense?
Maybe many of you are blockchain professionals but not familiar with cyber defense or blockchain for cyber defense.
Our main discussions are challenges in cyber defense, and then we will talk about related military projection analysis.
As a summary, we will talk about takeaway.
Targets of this talk are four kinds of people.
Someone interested in blockchain, someone interested in cyber defense, and someone interested in how blockchain can be hacked, and someone who needs fresh ideas.
Who doesn't know blockchain or what is cyber defense?
I hope this talk will give you fresh ideas about this topic or any ideas.
So blockchain for cyber defense, what is it?
Maybe you know what is blockchain.
Actually, blockchain is not clearly defined.
Blockchain is started from Bitcoin founded by Satoshi Nakamoto.
In his white paper, he didn't write clear definition of blockchain, but his main topic was Bitcoin.
In my thought, I think blockchain is constructed by three concepts.
Maybe some can think differently because the start of blockchain was not clearly defined.
One is, as a data structure, hash chain blocks, as a network structure, distributed network, and as a consensus structure, decentralized consensus.
And it has some fundamental problems.
It's called trilemma.
Scalability, security, and decentralization.
Three kinds of characteristics can be satisfied at the same time.
Then what is cyber defense?
Maybe some of you know, but maybe majority of you don't know.
This picture is just for fun.
It's not like cyber defense.
It's just funny.
So what is cyber defense?
Before I talk about this, I should define what is cyber security.
Classically, many people talk about information security, but also many people just talk about cyber security like information security.
If cyber security is the same as information security, then it's not worth to talk about cyber security or name cyber security.
So I bring a clear and distinguishable definition from the reference at the bottom you can see.
In their research, they define cyber security as treating stress using ICT.
Then what is classical information security?
Information security is treating stress to information-based assets.
But cyber security is treating any stress using ICT.
ICT is information and communication technology.
For example, car accident is not included in information security.
But if car accident is by a car remotely controlled by a hacker, then it is included in cyber security domain.
Now we can define what is cyber defense.
Cyber defense is intersection of cyber security and national security or national defense.
National security and national defense are used in the same way many times.
Not only cyber defense is about military defense, but widely it's all about national security.
So why is it like that?
You can see in the picture.
At the past, the stress to countries was through only military domain.
They should overcome armed forces, army, air force, and navy.
So general physical stress on defense is through military area.
But non-military domain, including severe assets and critical infrastructures, are covered by military domain.
But these days, they can be attacked directly by ICT, information and communication technologies.
For example, they can hack the plant and hack the financial systems and hack governmental information systems.
It's already happened.
In 2007, cyber attack to Estonia was really severe.
That's the beginning of the CCTCOE in NATO.
In 2010, Stuxnet to Iranian nuclear facilities.
It was shocking accident that computer virus can stop the nuclear targets and stop nuclear plant.
In 2015, Black Energy temporarily stopped the Ukrainian power grid.
Now, cyber attacks can directly attack national security and national assets.
Therefore, national countries need powerful cybersecurity technologies.
That's how cyber defense meets blockchain.
Blockchain looks secure and reliable.
Many articles and news always talk about it's really nice secure world with blockchain.
Secure, reliable, survival, and so on.
It's true.
Blockchain technology is not about single point of failure and it's immune to any minority attacks.
Even DDoS attacks are really difficult to blockchain.
It should overcome majority resources.
So, policy makers of countries easily think like that.
Defense systems require high security and high reliability.
Blockchain looks unhackable.
Okay, then defense needs this technology.
It's really simple logic and expectable.
In reality, many projects are already in progress.
We will discuss about related projects in the next section.
But in preview, military-included messaging app built on blockchain.
Blockchain supply chain enhancement for trusted and assured FPGA and ASICs.
Decentralized key management using blockchain.
Army innovation network provenance using blockchain and disconnected networks.
Navy multi-factor authentication.
Coups through resilient blockchain framework.
Chinese soldiers reward system using cryptocurrency.
Blockchain-based record intelligence.
French military police record on Tazos blockchain.
South Korean military.
Blockchain-based digital identification, DID.
And also other projects.
So, will it be as good as you think?
Well, in this talk, I want to question and introduce some challenges.
So, to meet the challenges, let's go to the battleground.
Maybe some of you like this game.
So, what is waiting for us in the battleground?
There is a profession. I introduced a profession, Carl von Clausewitz.
And he is a really professional in war and battleground.
He said in Twitter, war is a volume of uncertainty.
Three quarters of the factors on which action in war is based are wrapped in a fog of greater and lesser uncertainty.
In summary, he says, it's full of uncertainty.
So, what is big uncertainty?
Blockchain technology already assumes many uncertain situations, right?
But in battleground, uncertainty can be really extreme.
For example, you got ordered to go to mountain and save the mountains.
In Korea, mountains are like that.
And our territory is covered by mountains like that.
So, when you got ordered to go to mountain, then we expect mountains are covered by trees, usually, right?
But in reality, mountains can be devastated.
You can go to mountain and there can be even no single tree.
Rare trees burned down a year ago.
In Korean war in 1950, most of our trees burned down.
So, what I want to say is situations in battleground can be extremely miserable.
But you got ordered and you should say, I swear in army.
So, let's assume policy makers said adopt the blockchain technology and secure our networks.
Also, let's assume it's not questionable.
You are not allowed to question about this decision.
Actually, many people don't agree with private blockchain networks.
So, we will encounter three challenges.
Challenge one, air gap networks.
And challenge two, first dynamic environment.
And challenge three, the resource shortage.
So, challenge one, air gap networks.
So, what is air gap networks?
This is maybe familiar for most of you.
In military networks and critical infrastructures in most countries, many networks have air gap environments.
Air gap is something technology to make networks isolated.
So, you can see internet, mainly stressors come from outer internet.
So, to protect internet, you can make air gaps.
Air gaps can be software-based air gap and hardware-based air gap.
And there is true air gap.
True air gap means there is no communication between air gap.
But, well, it's not always magical.
Because Stuxnet exploded air gap networks.
Anyway, if there is an air gap, then it is hard to construct blockchain.
Because communication is not possible and many networks are partitioned and isolated.
So, anyway, let's make blockchain here.
So, for example, you can construct servers of each isolated network and communicate messages by those servers.
Then, it will be a single point of failure and it's just centralized.
Another way, you can make isolated networks and each isolated network has isolated blockchain.
But isolated networks have not enough nodes or servers to make enough decentralized networks.
So, it will make decentralization not effective.
So, in summary, because of the air gap structure of defense network, it may be hard to adopt blockchain.
And second challenge, first dynamic environment.
Wherever the military goes, communication must always exist.
Actually, it was same as the past warfare, but in modern warfare, getting information and process the information and make decisions are really important and should be fast.
Because we should make decisions faster than enemy and move faster than enemy.
So that we can make the superior position.
So, anyway, for this process, communication must be followed.
So, in this picture, in modern warfare, all of these things require communication.
So, it means, let's see the example.
It's map of World War II and second map is how Nazi extended their occupied territory really fast and broadly.
And right map is how their occupied territory shrinked really fast.
Let's see the left map.
When Nazi extended their occupied territory, then communication system must follow all of occupied territory and cover their communication really fast.
Even if their advance is really fast, communication and their system must follow. There's no exception.
At the right map, even if their occupied territory shrinks really fast, communication system will shrink really fast.
Anyway, just whatever is distributed throughout.
Another example. It's our country case. It's map of Korean War in 1950.
Before the Korean War, the territory was divided at this line, 38th parallel.
In three months, South Korea's territory was shrinked to just two cities, Daegu and Busan.
And we are really shrinking really fast.
But you can see, just in two or three months, we expanded our territory till the nearly end of North Korea.
But we retreated really fast again.
So, what I want to say is, better ground environment is really dynamic and we can't choose and there's no exception.
We are just first to meet situations.
So, let's move on the blockchain.
So, let's consider the case, sudden expansion.
Our private blockchain network is just like PBFT.
At the left situation, we have 8 total nodes and acceptable 30 nodes are like 2.
But we expanded our territory really fast and we just added a lot of nodes because we should cover the occupied territory.
Then, acceptable 30 nodes should be 12.
But in PBFT, this kind of consensus mechanism, we need settings.
But if this expansion is really fast and if this setting is not adapted, then it will be really easy to make consensus.
So, past majority becomes minority. 8 is minority of 36, so that is now your minority can make consensus.
Another case is kind of a reverse case.
So, let's think about sudden sharing key.
Sharing key is this typo.
Your original set is total nodes 36 and in PBFT, the number of acceptable 30 nodes is 11.
But most of nodes are just destroyed.
And then you have total nodes 8 and the number of acceptable 30 nodes should be 3.
But you are just destroyed, so you can't make setting in the time.
So, it will be nearly impossible to make consensus.
So, the past minority becomes the majority in this situation.
That is, now your majority can't make the consensus.
So, another situation, bombing and partitioning.
It's the last situation.
Your nodes just can't be partitioned.
In public blockchain, it's nearly impossible because nodes are connected by a worldwide web.
So, it's really hard to make clear partition.
Just divide into two partitions.
But it's possible in VR situations.
So, for example, you have to turn 36 nodes and partition them.
There's no majority anymore.
Furthermore, there's another issue.
If you connect the partitions but they have their own blocks,
especially for PVF tier deterministic consensus mechanisms, consistency will be a really big problem.
So, we'll meet our first problem.
In military situation, we can't just throw another information.
We can't choose just one information.
We need to collect all information and process and decide.
So, it'll be a problem.
So, it was the second challenge.
To summarize, in more extreme situations, assumptions can be easily broken.
Weakness can be easily revealed.
So, especially in my examples, deterministic consensus mechanisms that cannot guarantee aliveness are real problems in these situations.
Challenge 3, research shortage.
Actually, it's a classical problem of blockchain.
But I can't mention this problem.
Blockchain's other name is state replication system.
So, it'll replicate computation resource, storage resource, latency resource, bandwidth resource.
But existing military system were considered existing missions and existing systems.
So, just replicating existing performances will be a problem.
And we adopt blockchain for missions, not missions for blockchain.
So, influence on mission-critical functions should be checked.
Mission-critical functions should be guaranteed and then we can adopt blockchain.
But, yeah, first problem is resource consumption.
Then, what is second problem?
Second problem is serving resource consumption is not easy.
So, if we don't have enough resources, then we should get more resources and then we need to get more systems or more resources.
Then, we need to use defense acquisition process.
But maybe some of you already knew that military acquisition process is really conservative and it requires really rigorous tasks and evaluation process.
It's really heavy.
Defense acquisition process was having developed in this way.
So, you can see in the left picture, it's the US military acquisition system and it's really bureaucratic and conservative and really heavy.
So, there are many articles and interviews that DOD or many countries are suffering from full defense acquisition process.
Well, rigorous tasks and evaluations and rigorous acquisition system process is essential.
But, it is a really big problem in cybersecurity because, you know, system is developed and vulnerabilities are always hiding.
Vulnerabilities are just found time by time, but military acquisition system is not that rapid to react from these updates or kind of that.
So, anyway, military bureaucratic acquisition system process, because of that, it will be difficult to solve resource consumption problem.
So, in this situation, what can you do?
To adopt proof-of-work, we can't waste energy. Of course, proof-of-stake, we don't have coins in defense system.
And when we go to private consensus mechanism like PVRT, then we will go back to the challenge 2 again.
It can't guarantee liveness, so it will make such extreme problems.
So, in summary, mission is always first, but due to the military environment, support is not so timely and sufficient to adopt blockchain.
So, next is related military projects.
So, we surveyed 42 defense projects about blockchain and categorized as below.
And each project can be categorized into multiple categories.
For example, there's a research about key management about military drone system.
Then it's categorized into internet of things and communications and identification and authentication.
And especially data integrity category is just for data integrity project.
It's because adopting blockchain technology is basically about data integrity.
So, we categorized a project only about data integrity into that category.
So, main categories are like that. Data integrity, supply chain management, internet of things and communications, and identification and authentication.
So, we are somewhat careful about mentioning a project because about related project, detailed information about many real military projects are classified.
We got information from project notification, interviews, and news articles.
So, the information we got can be not clear or can be somewhat misunderstood.
So, we want to notice that our comments are based on limited information.
So, data integrity. For example, there are three projects.
French military police records on Taz's blockchain.
It's interesting that they put records on public blockchain.
So, maybe they can avoid problems that I mentioned like the problems happening in private blockchain.
And another example is U.S. DoD project sharing of defense research development testing evaluation on blockchain.
And another project is also U.S. DoD provenance using blockchain on disconnected networks.
This project is also interesting that they already considered some disconnected network situation in tactical networks.
So, another related category is supply chain management.
In cyber defense, supply chain management is really a big issue.
Maybe you read Bloomberg article about microchip hacking.
I don't know if it's true or not, but what I know is it was truly a big issue.
And there are other articles about supply chain attacks are suspicious.
Yeah, it's truly a really big, really important attack factor.
And so, there's a U.S. DoD project about it, blockchain supply chain enhancements for trusted and assured LPGA and ASICs.
But blockchain on supply management is actively researched for efficiency about supply chain, not only for security.
Another category is Internet of Things.
There's a U.S. Department of Homeland Security project combining blockchain technology with critical infrastructure.
This project is about sensors and cameras that protect integrity and authenticity of critical infrastructure.
This project is about IoT, so we can just relate it to Challenge 3 research series.
But I think the infrastructure is not dynamic, so it can be adaptable, I think.
So, in communication category, there's a DARPA project building an encrypted message system based on blockchain technology.
I think it's about tactical networks, so I don't know the detailed information, but if it is related to messages in tactical networks.
Challenge 2, first dynamic environments, and Challenge 3 research series should be considered. Maybe they will be considered.
For identification and authentication, there are two projects.
For example, a U.S. DHS project decentralizes KME management using blockchain.
There's a South Korea project blockchain-based DID.
But maybe some of you noticed, some of these projects are not domain-specific.
I mean, it's not just only for cyber defense. Many projects are just general.
But for this project, especially identification and authentication, the decentralized key management and DID is not only military issue, not only cyber defense.
So maybe they are not interesting.
So, related projects in my examples are like that.
Lines mean that projects are not domain-specific.
I'm interested in blockchain specifically for cyber defense.
So I'm still surveying about this topic and I'm preparing a paper.
Actually, I think I criticized a lot with just question this topic and introduced challenges.
But then some can ask, then what are alternatives?
What I want to say is we don't need to choose a perfect blockchain structure.
As I mentioned at the first part of this talk, blockchain is constructed by three concepts.
As a data structure, hash chain blocks.
And as a network structure, distributed network.
And as a consensus structure, decentralized consensus.
But you can just give up one concept or two concepts.
Maybe you only need distributed network and penalized networks.
So what I need to do is figure out problems and figure out right solutions.
Not just decide to use blockchain and figure out which blockchain is adaptable.
So to summarize, takeaway are three.
Cyber defense makes more challenges and requirements for blockchain.
Error gaps, sudden expansion, shrinkage, partitioning.
Extreme situation can exist.
Second, blockchain's resource consumption can be a problem.
But defense environments are not that flexible to assign more resources.
Third, we don't need to cling to blockchain if it's not adaptable.
Otherwise, it will bring more issues.
So it's all of my talk.
And get in touch to these emails.
We welcome any questions and discussions.
And I think this presentation will be uploaded somewhere so you can see our emails.
So thank you for your attention and thank you again.
Any questions?
Is there a question?
If you can't ask now, then you can just send email or send message to Discord.
Any questions and discussions are welcome.
Okay, then. Thank you.
Then I can just quit the live streaming, right?