Security and Privacy in an Edge Analytics World

When deploying an application to the internet, the risks posed by cyber criminals should be taken very seriously. Internet-enabled devices, including edge computers, are prone to cyber attacks, where they may be used to shut down websites or cause havoc on the internet—not to mention the destruction that cyber attacks can cause to our networked applications.

In this chapter, we will cover security—and, in turn, privacy—when it comes to our edge analytics applications.

The following topics will be covered in this chapter:

• An overview of the Internet-of-Things security
• Types of attacks against our edge analytics applications
• Protecting our edge analytics applications
• Monitoring and auditing our edge analytics applications

An overview of the Internet-of-Things security

Starting in September 2016, the Mirai botnet launched a distributed denial of service (DDoS) attack using the Internet of Things (IoT) devices on some of the biggest websites in the world, including Reddit, Airbnb, and Netflix. Flooding the DNS provider of these websites with more than 1 TB of data per second from captured devices, the cyber attack was one of the biggest DDoS attacks ever recorded.

So, what exactly is a DDoS attack and how are IoT devices involved? The following diagram shows what a DDoS attack with IoT devices might look like:

DDoS is an attack on a website that overwhelms it with excessive traffic, causing the website to fail. This type of attack is beneficial to the website's competitors. In the case of the Mirai botnet, the attack was designed by students interested in disabling competing Minecraft servers...

Types of attacks against our edge analytics applications

By its very nature, an edge analytics application is more secure than a pure IoT application. This is due to the fact that we limit the amount of information that is transmitted over the internet. However, there are still security concerns we must take into account.

The following are three examples of these concerns.

Vulnerability issues

Vulnerabilities in our internet-facing interfaces may be exploited by cyber criminals. These vulnerabilities include using default passwords and out-of-date software updates. In the following diagram, we can see the vulnerabilities of the internet-facing interfaces of the smart doorbell application we built in the previous chapter:

...

Protecting our edge analytics applications

Protecting our edge analytics applications may be a simple matter of changing our passwords frequently or placing our edge devices in secure locations. Also, the use of SSL certificates ensures that the data we transmit across the internet is encrypted. In this section, we will discuss some of the methods of ensuring our edge analytics applications are as secure as possible from cyber crime. We will start by looking at passwords.

Passwords and updates

In Chapter 6, Using MicroPython for Edge Analytics, we used a few microcontrollers as edge devices. Microcontrollers, such as Pycom's LoPy, use default usernames and passwords for both the WiFi access point and FTP file transfers...

Monitoring and auditing our edge analytics applications

Security is a big concern for our applications and must be taken into account when building and deploying them. The question we should ask is how do we get on top of it? The key to mitigating security concerns is monitoring and auditing the devices on our applications. By analyzing the traffic that flows into our applications, we can set up alerts to let us know whether there are any possible security threats. Also, by keeping an audit of our devices and addressing possible security concerns before deployment, we can patch known security holes before they become a problem.

Let's take a look at monitoring, first.

Monitoring our edge analytics applications

For edge...

Summary

In this chapter, we started our discussion on security and privacy by describing a cyber attack involving IoT devices. Although our edge analytics applications offer a bit of a buffer against the unauthorized use of devices, there is still a lot that should concern us when it comes to security.

We then looked at some of the various types of attacks that we should look out for when building and deploying our applications. These include issues with vulnerability, sniffing of our data, and spoofing of our devices, such that our application confuses the fake device for one of our own.

We then took a look at how we can protect our edge analytics applications by confirming the updates of passwords and security patches. We discussed how physical security can be implemented to keep remote devices safe. We also looked at how secure internet connections work and took a high-level...

Questions

Having read this chapter, try answering the following questions on your own:

1. True/False. The code used in the Mirai botnet attack in 2016 was released to the public in order to create plausible deniability for the creators.
2. True/False. A command and control server is used in DDoS attacks.
3. What is a cross-site scripting attack?
4. Name three of the possible types of attacks against our edge analytics applications.
5. True/False. A sniffer is a type of spying tool that is set up between our internet router and edge computer.
6. How does a cyber criminal trick you into giving them your username and password when you log on to a legitimate website?
7. True/False. A physical key switch can be used as a form of security protection on our edge device.
8. What are some of the libraries that we can use to write a machine learning module?
9. True/False. The name of the Microsoft tool used to monitor...

Further reading

For more information about Azure Security Center for IoT, go to https://docs.microsoft.com/en-us/azure/asc-for-iot/overview.