Hunting the Ghost: Towards Automatic Mining of IoT Hidden Services

Shuaike Dong, Siyu Shen, Zhou Li, Kehuan Zhang · 2025

Read Hunting the Ghost: Towards Automatic Mining of IoT Hidden Services by Shuaike Dong, Siyu Shen, Zhou Li, Kehuan Zhang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-06822.txt.

Introduction

Shuaike Dong, Siyu Shen, Zhou Li, and Kehuan Zhang present a 2025 paper on a tool that hunts hidden services inside the software of smart devices. A hidden service runs on a device without the owner's knowledge, yet an attacker who finds it can use it as a door. Finding such services by hand is slow.

Their tool scans device firmware with static analysis and symbolic execution. It filters for services that stay invisible to normal users but stand out to a skilled attacker. A prototype was tested on a set of real firmware images. The authors report that it found the suspicious hidden services well.

Catalog record

CREATORShuaike Dong, Siyu Shen, Zhou Li, Kehuan Zhang
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE44 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-06822

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Hunting the Ghost: Towards Automatic Mining of IoT Hidden Services Shuaike Dong, Siyu Shen, Zhou Li, Kehuan Zhang Abstract The Internet of Things (IoT) has been prosperously used and profoundly changing people’s daily life. However, for normal users, most IoT devices are smart “black boxes” listening to their commands and providing feedbacks. Recent notorious IoT-based attacks have shown that even IoT devices themselves may behave normally, evil things can be happening under the hood. Therefore, it is crucial for both normal users and security analysts to quickly profile the behavior of an IoT device and dig out ”open” and ”half-open” doors in it. In this paper, we proposes IoTBolt , an automatic firmware analysis tool targeting at finding hidden services that may be potentially harmful to the IoT devices. Our approach uses static analysis and symbolic execution to search and filter services that are transparent to normal users but explicit to experienced attackers. A prototype is built and evaluated against a dataset of IoT firmware, and The evaluation shows our tool can find the suspicious hidden services effectively. PubID: pubid: Network and Distributed Systems Security (NDSS) Symposium 2020 23-26 February 2020, San Diego, CA, USA ISBN 1-891562-61-4 https://dx.doi.org/10.14722/ndss.2020.23xxx www.ndss-symposium.org I Background I-A Smart-Home IoT Architecture A typical smart home consists of three parties: IoT device, service provider and user. The normal interaction scheme is shown in fig . Below we briefly introduce it. Fig. 1: A typical smart home I-B IoT Firmware IoT firmware is the software programmed on IoT hardware. Since IoT products are usually embedded devices with limited computation resources and variant architectures. IoT firmware provides basic functionalities to interact with the hardware. Firmware layout. There are two typical types of firmware, firmware blobs and embedded-Linux based firmware . Firmware blobs often come with bare-metal devices with very monotonous capabilities and restricted resources, like smart bulbs and smart plugs. Firmware blobs usually have various formats due to the proprietary development process of their vendors. According to previous works [ 22 ] , embedded-Linux based firmware takes up the largest portion of off-the-shelf firmware with a ratio of around 86%. Contents of an embedded-Linux based firmware vary among different vendors. In most cases, there is at least one embedded file system inside and can be extracted by firmware unpacking tools like binwalk [ 14 ] . Firmware acquisition. There are generally three ways to obtain a firmware. 1) Download from websites of the vendor. Several vendors release the firmware dataset on their websites, like D-Link [ 5 ] and TP-Link [ 11 ] . 2) Extract firmware from hardware. Firmware often resides in the storage unit of an IoT device such as flash chip. With microcontrollers that supports flash communication protocols, one can dump the firmware without destroying the device [ 9 ] . Apart from flash chip dumping, other hardware debugging ports can be leveraged for the firmware extraction as well [ 8 ] . 3) Network interception. Some vendors provide mobile Apps to facilitate the control of IoT devices. Most of these Apps can trigger the firmware upgrade by one or several simple operations. During the firmware upgrading process, either the meta-data (including download url, file size and checksum, etc.) or the firmware itself will be transmitted to the device. An experienced analyst can easily intercept the traffic he/she needs by setting up a proxy. I-C Issues in IoT Development As the development of a commercial IoT product is a long and lasting task during which the code may experience a lot of changes and updates from different developers. Many issues may arise during the process and thus produce some ”hidden” services, which may be utilized by malicious attackers. We classify the factors to these risks as follows: 1. Leftover Debugging code. As the modern software development is usually a complicated process including multiple coupling components, developers widely use debugging to locate the errors and fix them. Different from codes released in production version, debugging code usually involves a wider capability and higher privilege, or some functionalities that are just for testing usage. Apart from that, the debugging codes are usually not directly visible to normal users but can be easily found by experienced and purposeful attackers. 2. Unmatched interfaces and services. To facilitate normal users in using IoT devices, vendors usually provide easy-to-use interfaces for them. For example, most domestic routers provide a series of web pages leading users to configure the product step by step. As those pages are just interactive entrances to back-end running programs, we call them interfaces and their targets as services . From the perspective of a normal user, interfaces and services should have a ”surjection” relation which means each service should have at least one interface to trigger it. However, there are cases when multiple services running behind the scene without any companion interfaces. Apart from that, even those ”already-matched” services may contain hidden functions that are not within the scope of their interfaces, which are exploitable by experienced attackers. 3. Backdoors. Previous work [ 25 ] has shown that backdoors are pervasive in IoT devices. From the perspective of vendors, backdoors are helpful for debugging usage and testing phase during the development. However, they can be taken advantage of by attackers and thus leading to the compromise of the whole device or privacy exposure. I-D IoT interfaces and services As the medium between normal users and the environment, IoT devices provide easy-to-use interfaces for consumers. We categorize common interfaces as follows: 1. Web-based interfaces. Some IoT devices embed web pages inside the firmware to help users do the configuration. For example, home-use routers often host the domain “http://192.168.0.1” by a web server. People access that domain and set up the router step by step according to its wizard. According to our study, some IoT cameras, also support such kind of interface. 2. Mobile-based interfaces. More and more IoT developers are releasing their companion mobile Apps with devices. A user yields commands by operating the UI controllers of the mobile App, the commands will be transferred to the cloud or directly sent to the device in the form of network packets. 3. Physical contacts. Most IoT devices provide physical interfaces to make them consistent with previous non-IoT devices, like traditional plugs and sockets. Some physical actions like long-pressing a button also act as the critical role in the binding process, like [ 21 ] shows. 4. Others. There are additional interfaces utilizing different kinds of signals. For example, Google home [ 15 ] responds when it hears legal voice commands. Some smart thermometers commits certain behaviors when the temperature has reached a threshold. After a user request is well received by the device, it is processed by the software components of the IoT firmware. Due to the variety of the information involved inside requests, different handlers are chosen to react. We refer to reactions taken by the IoT devices as services , which are IoT functionalities that can be aroused through IoT interfaces. To conclude, IoT services hold the following attributes: 1. can be triggered through IoT interfaces or channels with the equal semantics. For instance, a service of a home router can be activated either by interacting with the

Read the complete text