Reverse engineering (also known as backwards engineering or back engineering) is a process or method through which one attempts to understand through deductive reasoning how a previously made device, process, system, or piece of software accomplishes a task with very little (if any) insight into exactly how it does so. It is essentially the process of opening up or dissecting a system to see how it works, in order to duplicate or enhance it. Depending on the system under consideration and the technologies employed, the knowledge gained during reverse engineering can help with repurposing obsolete objects, doing security analysis, or learning how something works.
Although the process is specific to the object on which it is being performed, all reverse engineering processes consist of three basic steps: Information extraction, Modeling, and Review. Information extraction refers to the practice of gathering all relevant information for performing the operation. Modeling refers to the practice of combining the gathered information into an abstract model, which can be used as a guide for designing the new object or system. Review refers to the testing of the model to ensure the validity of the chosen abstract. Reverse engineering is applicable in the fields of computer engineering, mechanical engineering, design, electronic engineering, software engineering, chemical engineering, and systems biology.
There are many reasons for performing reverse engineering in various fields. Reverse engineering has its origins in the analysis of hardware for commercial or military advantage.: 13 However, the reverse engineering process may not always be concerned with creating a copy or changing the artifact in some way. It may be used as part of an analysis to deduce design features from products with little or no additional knowledge about the procedures involved in their original production.: 15
In some cases, the goal of the reverse engineering process can simply be a redocumentation of legacy systems.: 15  Even when the reverse-engineered product is that of a competitor, the goal may not be to copy it but to perform competitor analysis. Reverse engineering may also be used to create interoperable products and despite some narrowly-tailored United States and European Union legislation, the legality of using specific reverse engineering techniques for that purpose has been hotly contested in courts worldwide for more than two decades.
Software reverse engineering can help to improve the understanding of the underlying source code for the maintenance and improvement of the software, relevant information can be extracted to make a decision for software development and graphical representations of the code can provide alternate views regarding the source code, which can help to detect and fix a software bug or vulnerability. Frequently, as some software develops, its design information and improvements are often lost over time, but that lost information can usually be recovered with reverse engineering. The process can also help to cut down the time required to understand the source code, thus reducing the overall cost of the software development. Reverse engineering can also help to detect and to eliminate a malicious code written to the software with better code detectors. Reversing a source code can be used to find alternate uses of the source code, such as detecting the unauthorized replication of the source code where it was not intended to be used, or revealing how a competitor's product was built. That process is commonly used for "cracking" software and media to remove their copy protection,: 7 or to create a possibly-improved copy or even a knockoff, which is usually the goal of a competitor or a hacker.: 8
Malware developers often use reverse engineering techniques to find vulnerabilities in an operating system to build a computer virus that can exploit the system vulnerabilities.: 5 Reverse engineering is also being used in cryptanalysis to find vulnerabilities in substitution cipher, symmetric-key algorithm or public-key cryptography.: 6
There are other uses to reverse engineering:
As computer-aided design (CAD) has become more popular, reverse engineering has become a viable method to create a 3D virtual model of an existing physical part for use in 3D CAD, CAM, CAE, or other software. The reverse-engineering process involves measuring an object and then reconstructing it as a 3D model. The physical object can be measured using 3D scanning technologies like CMMs, laser scanners, structured light digitizers, or industrial CT scanning (computed tomography). The measured data alone, usually represented as a point cloud, lacks topological information and design intent. The former may be recovered by converting the point cloud to a triangular-faced mesh. Reverse engineering aims to go beyond producing such a mesh and to recover the design intent in terms of simple analytical surfaces where appropriate (planes, cylinders, etc.) as well as possibly NURBS surfaces to produce a boundary-representation CAD model. Recovery of such a model allows a design to be modified to meet new requirements, a manufacturing plan to be generated, etc.
Hybrid modeling is a commonly used term when NURBS and parametric modeling are implemented together. Using a combination of geometric and freeform surfaces can provide a powerful method of 3D modeling. Areas of freeform data can be combined with exact geometric surfaces to create a hybrid model. A typical example of this would be the reverse engineering of a cylinder head, which includes freeform cast features, such as water jackets and high-tolerance machined areas.
Reverse engineering is also used by businesses to bring existing physical geometry into digital product development environments, to make a digital 3D record of their own products, or to assess competitors' products. It is used to analyze how a product works, what it does, what components it has; estimate costs; identify potential patent infringement; etc.
Value engineering, a related activity that is also used by businesses, involves deconstructing and analyzing products. However, the objective is to find opportunities for cost-cutting.
Main article: PCB reverse engineering
Reverse engineering of printed circuit boards involves recreating fabrication data for a particular circuit board. This is done to allow benchmarking, and support for legacy systems.
In 1990, the Institute of Electrical and Electronics Engineers (IEEE) defined (software) reverse engineering (SRE) as "the process of analyzing a subject system to identify the system's components and their interrelationships and to create representations of the system in another form or at a higher level of abstraction" in which the "subject system" is the end product of software development. Reverse engineering is a process of examination only, and the software system under consideration is not modified, which would otherwise be re-engineering or restructuring. Reverse engineering can be performed from any stage of the product cycle, not necessarily from the functional end product.
There are two components in reverse engineering: redocumentation and design recovery. Redocumentation is the creation of new representation of the computer code so that it is easier to understand. Meanwhile, design recovery is the use of deduction or reasoning from general knowledge or personal experience of the product to understand the product's functionality fully. It can also be seen as "going backwards through the development cycle". In this model, the output of the implementation phase (in source code form) is reverse-engineered back to the analysis phase, in an inversion of the traditional waterfall model. Another term for this technique is program comprehension. The Working Conference on Reverse Engineering (WCRE) has been held yearly to explore and expand the techniques of reverse engineering. Computer-aided software engineering (CASE) and automated code generation have contributed greatly in the field of reverse engineering.
Software anti-tamper technology like obfuscation is used to deter both reverse engineering and re-engineering of proprietary software and software-powered systems. In practice, two main types of reverse engineering emerge. In the first case, source code is already available for the software, but higher-level aspects of the program, which are perhaps poorly documented or documented but no longer valid, are discovered. In the second case, there is no source code available for the software, and any efforts towards discovering one possible source code for the software are regarded as reverse engineering. The second usage of the term is more familiar to most people. Reverse engineering of software can make use of the clean room design technique to avoid copyright infringement.
On a related note, black box testing in software engineering has a lot in common with reverse engineering. The tester usually has the API but has the goals to find bugs and undocumented features by bashing the product from outside.
Other purposes of reverse engineering include security auditing, removal of copy protection ("cracking"), circumvention of access restrictions often present in consumer electronics, customization of embedded systems (such as engine management systems), in-house repairs or retrofits, enabling of additional features on low-cost "crippled" hardware (such as some graphics card chip-sets), or even mere satisfaction of curiosity.
Binary reverse engineering is performed if source code for a software is unavailable. This process is sometimes termed reverse code engineering, or RCE. For example, decompilation of binaries for the Java platform can be accomplished by using Jad. One famous case of reverse engineering was the first non-IBM implementation of the PC BIOS, which launched the historic IBM PC compatible industry that has been the overwhelmingly-dominant computer hardware platform for many years. Reverse engineering of software is protected in the US by the fair use exception in copyright law. The Samba software, which allows systems that do not run Microsoft Windows systems to share files with systems that run it, is a classic example of software reverse engineering since the Samba project had to reverse-engineer unpublished information about how Windows file sharing worked so that non-Windows computers could emulate it. The Wine project does the same thing for the Windows API, and OpenOffice.org is one party doing that for the Microsoft Office file formats. The ReactOS project is even more ambitious in its goals by striving to provide binary (ABI and API) compatibility with the current Windows operating systems of the NT branch, which allows software and drivers written for Windows to run on a clean-room reverse-engineered free software (GPL) counterpart. WindowsSCOPE allows for reverse-engineering the full contents of a Windows system's live memory including a binary-level, graphical reverse engineering of all running processes.
Another classic, if not well-known, example is that in 1987 Bell Laboratories reverse-engineered the Mac OS System 4.1, originally running on the Apple Macintosh SE, so that it could run it on RISC machines of their own.
Reverse engineering of software can be accomplished by various methods. The three main groups of software reverse engineering are
Software classification is the process of identifying similarities between different software binaries (such as two different versions of the same binary) used to detect code relations between software samples. The task was traditionally done manually for several reasons (such as patch analysis for vulnerability detection and copyright infringement), but it can now be done somewhat automatically for large numbers of samples.
This method is being used mostly for long and thorough reverse engineering tasks (complete analysis of a complex algorithm or big piece of software). In general, statistical classification is considered to be a hard problem, which is also true for software classification, and so few solutions/tools that handle this task well.
A number of UML tools refer to the process of importing and analysing source code to generate UML diagrams as "reverse engineering". See List of UML tools.
Although UML is one approach in providing "reverse engineering" more recent advances in international standards activities have resulted in the development of the Knowledge Discovery Metamodel (KDM). The standard delivers an ontology for the intermediate (or abstracted) representation of programming language constructs and their interrelationships. An Object Management Group standard (on its way to becoming an ISO standard as well), KDM has started to take hold in industry with the development of tools and analysis environments that can deliver the extraction and analysis of source, binary, and byte code. For source code analysis, KDM's granular standards' architecture enables the extraction of software system flows (data, control, and call maps), architectures, and business layer knowledge (rules, terms, and process). The standard enables the use of a common data format (XMI) enabling the correlation of the various layers of system knowledge for either detailed analysis (such as root cause, impact) or derived analysis (such as business process extraction). Although efforts to represent language constructs can be never-ending because of the number of languages, the continuous evolution of software languages, and the development of new languages, the standard does allow for the use of extensions to support the broad language set as well as evolution. KDM is compatible with UML, BPMN, RDF, and other standards enabling migration into other environments and thus leverage system knowledge for efforts such as software system transformation and enterprise business layer analysis.
Protocols are sets of rules that describe message formats and how messages are exchanged: the protocol state machine. Accordingly, the problem of protocol reverse-engineering can be partitioned into two subproblems: message format and state-machine reverse-engineering.
The message formats have traditionally been reverse-engineered by a tedious manual process, which involved analysis of how protocol implementations process messages, but recent research proposed a number of automatic solutions. Typically, the automatic approaches group observe messages into clusters by using various clustering analyses, or they emulate the protocol implementation tracing the message processing.
There has been less work on reverse-engineering of state-machines of protocols. In general, the protocol state-machines can be learned either through a process of offline learning, which passively observes communication and attempts to build the most general state-machine accepting all observed sequences of messages, and online learning, which allows interactive generation of probing sequences of messages and listening to responses to those probing sequences. In general, offline learning of small state-machines is known to be NP-complete, but online learning can be done in polynomial time. An automatic offline approach has been demonstrated by Comparetti et al. and an online approach by Cho et al.
Other components of typical protocols, like encryption and hash functions, can be reverse-engineered automatically as well. Typically, the automatic approaches trace the execution of protocol implementations and try to detect buffers in memory holding unencrypted packets.
Reverse engineering is an invasive and destructive form of analyzing a smart card. The attacker uses chemicals to etch away layer after layer of the smart card and takes pictures with a scanning electron microscope (SEM). That technique can reveal the complete hardware and software part of the smart card. The major problem for the attacker is to bring everything into the right order to find out how everything works. The makers of the card try to hide keys and operations by mixing up memory positions, such as by bus scrambling.
In some cases, it is even possible to attach a probe to measure voltages while the smart card is still operational. The makers of the card employ sensors to detect and prevent that attack. That attack is not very common because it requires both a large investment in effort and special equipment that is generally available only to large chip manufacturers. Furthermore, the payoff from this attack is low since other security techniques are often used such as shadow accounts. It is still uncertain whether attacks against chip-and-PIN cards to replicate encryption data and then to crack PINs would provide a cost-effective attack on multifactor authentication.
Full reverse engineering proceeds in several major steps.
The first step after images have been taken with a SEM is stitching the images together, which is necessary because each layer cannot be captured by a single shot. A SEM needs to sweep across the area of the circuit and take several hundred images to cover the entire layer. Image stitching takes as input several hundred pictures and outputs a single properly-overlapped picture of the complete layer.
Next, the stitched layers need to be aligned because the sample, after etching, cannot be put into the exact same position relative to the SEM each time. Therefore, the stitched versions will not overlap in the correct fashion, as on the real circuit. Usually, three corresponding points are selected, and a transformation applied on the basis of that.
To extract the circuit structure, the aligned, stitched images need to be segmented, which highlights the important circuitry and separates it from the uninteresting background and insulating materials.
Finally, the wires can be traced from one layer to the next, and the netlist of the circuit, which contains all of the circuit's information, can be reconstructed.
Reverse engineering is often used by people to copy other nations' technologies, devices, or information that have been obtained by regular troops in the fields or by intelligence operations. It was often used during the Second World War and the Cold War. Here are well-known examples from the Second World War and later:
Reverse engineering concepts have been applied to biology as well, specifically to the task of understanding the structure and function of gene regulatory networks. They regulate almost every aspect of biological behavior and allow cells to carry out physiological processes and responses to perturbations. Understanding the structure and the dynamic behavior of gene networks is therefore one of the paramount challenges of systems biology, with immediate practical repercussions in several applications that are beyond basic research. There are several methods for reverse engineering gene regulatory networks by using molecular biology and data science methods. They have been generally divided into six classes:
Often, gene network reliability is tested by genetic perturbation experiments followed by dynamic modelling, based on the principle that removing one network node has predictable effects on the functioning of the remaining nodes of the network. Applications of the reverse engineering of gene networks range from understanding mechanisms of plant physiology to the highlighting of new targets for anticancer therapy.
Reverse engineering applies primarily to gaining understanding of a process or artifact in which the manner of its construction, use, or internal processes has not been made clear by its creator.
Patented items do not of themselves have to be reverse-engineered to be studied, for the essence of a patent is that inventors provide a detailed public disclosure themselves, and in return receive legal protection of the invention that is involved. However, an item produced under one or more patents could also include other technology that is not patented and not disclosed. Indeed, one common motivation of reverse engineering is to determine whether a competitor's product contains patent infringement or copyright infringement.
In the United States, even if an artifact or process is protected by trade secrets, reverse-engineering the artifact or process is often lawful if it has been legitimately obtained.
Reverse engineering of computer software often falls under both contract law as a breach of contract as well as any other relevant laws. That is because most end-user license agreements specifically prohibit it, and US courts have ruled that if such terms are present, they override the copyright law that expressly permits it (see Bowers v. Baystate Technologies). According to Section 103(f) of the Digital Millennium Copyright Act (17 U.S.C. § 1201 (f)), a person in legal possession of a program may reverse-engineer and circumvent its protection if that is necessary to achieve "interoperability", a term that broadly covers other devices and programs that can interact with it, make use of it, and to use and transfer data to and from it in useful ways. A limited exemption exists that allows the knowledge thus gained to be shared and used for interoperability purposes.[a]
EU Directive 2009/24 on the legal protection of computer programs, which superseded an earlier (1991) directive, governs reverse engineering in the European Union.[b]
The unauthorised reproduction, translation, adaptation or transformation of the form of the code in which a copy of a computer program has been made available constitutes an infringement of the exclusive rights of the author. Nevertheless, circumstances may exist when such a reproduction of the code and translation of its form are indispensable to obtain the necessary information to achieve the interoperability of an independently created program with other programs. It has therefore to be considered that, in these limited circumstances only, performance of the acts of reproduction and translation by or on behalf of a person having a right to use a copy of the program is legitimate and compatible with fair practice and must therefore be deemed not to require the authorisation of the rightholder. An objective of this exception is to make it possible to connect all components of a computer system, including those of different manufacturers, so that they can work together. Such an exception to the author's exclusive rights may not be used in a way which prejudices the legitimate interests of the rightholder or which conflicts with a normal exploitation of the program.