Cyber attack early warning system

Abstract

A system and method for generating an alert regarding a potential attack is described. The method involves receiving data associated with previously analyzed or known malware attacks by a first network device. Additionally, the first network device receives an attack alert associated with an object analyzed and identified as suspicious by a second network device. The attack alert includes information associated with the suspicious object. For alert generation, at least a portion of the information of the attack alert is provided to a system configured to at least (i) extract feature(s) from the attack alert, (ii) determine similarities between the extracted features and features associated with the previously analyzed or known malware attacks to determine a result, (iv) compute an attack value based on the result and at least a portion of the extracted features including time-dependent and/or independent features, and (v) generate an alert based on the attack value.

Description

FIELD

Embodiments of the disclosure relate to the field of cyber-security. More specifically, one embodiment of the disclosure relates to a system, apparatus and method for providing an early warning of a potential attack on a client device based on real-time analysis.

GENERAL BACKGROUND

Over the last decade, malicious software has become a pervasive problem for Internet users as many networked resources include vulnerabilities that are subject to attack. For instance, over the past few years, an increasing number of vulnerabilities are being discovered in software that is loaded onto network devices, such as vulnerabilities within operating systems, for example. While some vulnerabilities continue to be addressed through software patches, prior to the release of such software patches, network devices will continue to be targeted for attack by malware, namely information such as computer code that attempts during execution to take advantage of a vulnerability in computer software by acquiring sensitive information or adversely influencing or attacking normal operations of the network device or the entire enterprise network.

Moreover, with the proliferation of the Internet, social media, and email, malware is capable of spreading more quickly and effecting a larger subset of the population than ever before. In response, systems have been developed to remove and/or quarantine malware. For example, users are generally informed to “scan” their various devices and/or systems using traditional security measures such as “anti-virus” software that is configured to remove any malware that may have been installed therein. Unfortunately, users are at the mercy of the most recent update of their anti-virus software, which may or may not include information to prevent possible malware exploits.

Furthermore, modern malware is becoming increasingly targeted, such that particular users, industries, or even specific segments are being sought, thereby circumventing even the most recently updated and detailed security measures that are usually generic in nature. The issue is further exacerbated because malware is also becoming increasingly sophisticated, to the point that such malware may go undetected by anti-virus software, or go so far as circumventing anti-virus measures altogether.

Accordingly, a need exists for an early warning system that is preventative, rather than reactive after the fact, to predict potential malware attacks prior to their occurrence, and notify targets that have a high probability of being attacked.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention are illustrated by way of example and not by way of limitation in the figures of the accompanying drawings, in which like references indicate similar elements and in which:

FIG. 1 is an exemplary block diagram of an operational flow of an early warning system.

FIG. 2A is an exemplary block diagram of a communication system deploying a cyber attack early warning system via a network.

FIG. 2B is an exemplary embodiment of a logical representation of the cyber attack early warning system of FIG. 2A.

FIG. 3 is a flowchart of an exemplary method for providing early warning of a potential attack on a client device based on a real-time analysis of a plurality of factors.

FIG. 4 is an exemplary network device demonstrating an alert according to the present disclosure.

DETAILED DESCRIPTION

Embodiments of the present disclosure generally relate to a cyber attack early warning system that is configured to predict and notify probable targets of a potential malware attack prior to its occurrence.

According to one embodiment of the disclosure, network traffic is analyzed such that malicious network traffic is filtered and preferably prevented from executing, using a remote source. Herein, upon detection of potentially malicious network traffic, the remote source is configured to generate an “attack alert” that is communicated to the cyber attack early warning system, which comprises an input engine, an expert engine, a correlation engine, and a reporting engine.

When the attack alert is received by the cyber attack early warning system, the attack alert is analyzed and classified. For example, the attack alert may correspond to any of various malware attack types, including without limitation Advanced Persistent Threats (APT), Point-Of-Sales attacks (POS), Crimeware attacks, and the like.

Once the attack alert is classified, an attack-specific engine is configured to provide further in-depth analysis of the attack alert, including the application of a plurality of analysis mechanisms, such as various algorithms and/or models specific to the type of malware attack. It is contemplated that the attack-specific engine may also analyze elements of past attacks and related data. The results of these various analyses are correlated to compute an “attack value” so that probable attack targets may be notified. For example, probable attack targets may include companies or individuals in the same industry, geographic region, and capacity, for example, without limitation. If the determined attack value matches or exceeds a predetermined threshold value, then an early warning alert is provided to probable targets of a malware attack. If the attack value is less than the predetermined threshold, then the early warning alert is not generated.

I. Terminology

In the following description, certain terminology is used to describe features of the invention. For example, in certain situations, both terms “logic” and “engine” are representative of hardware, firmware and/or software that is configured to perform one or more functions. As hardware, logic (or engine) may include circuitry having data processing or storage functionality. Examples of such circuitry may include, but are not limited or restricted to a microprocessor, one or more processor cores, a programmable gate array, a microcontroller, an application specific integrated circuit, wireless receiver, transmitter and/or transceiver circuitry, semiconductor memory, or combinatorial logic.

Logic (or engine) may be software in the form of one or more software modules, such as executable code in the form of an executable application, an application programming interface (API), a subroutine, a function, a procedure, an applet, a servlet, a routine, source code, object code, a shared library/dynamic load library, or one or more instructions. These software modules may be stored in any type of a suitable non-transitory storage medium, or transitory storage medium (e.g., electrical, optical, acoustical or other form of propagated signals such as carrier waves, infrared signals, or digital signals). Examples of non-transitory storage medium may include, but are not limited or restricted to a programmable circuit; a semiconductor memory; non-persistent storage such as volatile memory (e.g., any type of random access memory “RAM”); persistent storage such as non-volatile memory (e.g., read-only memory “ROM”, power-backed RAM, flash memory, phase-change memory, etc.), a solid-state drive, hard disk drive, an optical disc drive, or a portable memory device. As firmware, the executable code is stored in persistent storage.

The term “object” generally refers to a collection of data, whether in transit (e.g., over a network) or at rest (e.g., stored), often having a logical structure or organization that enables it to be classified for purposes of analysis. During analysis, for example, the object may exhibit a set of expected characteristics and, during processing, a set of expected behaviors. The object may also exhibit a set of unexpected characteristics and a set of unexpected behaviors that may evidence an exploit and potentially allow the object to be classified as an exploit.

Examples of objects may include one or more flows or a self-contained element within a flow itself. A “flow” generally refers to related packets that are received, transmitted, or exchanged within a communication session. For convenience, a packet is broadly referred to as a series of bits or bytes having a prescribed format, which may include packets, frames, or cells.

As an illustrative example, an object may include a set of flows such as (1) a sequence of transmissions in accordance with a particular communication protocol (e.g., User Datagram Protocol (UDP); Transmission Control Protocol (TCP); or Hypertext Transfer Protocol (HTTP); etc.), or (2) inter-process communications (e.g., Remote Procedure Call “RPC” or analogous processes, etc.). Similar, as another illustrative example, the object may be a self-contained element, where different types of such objects may include an executable file, non-executable file (such as a document or a dynamically link library), a Portable Document Format (PDF) file, a JavaScript file, Zip file, a Flash file, a document (for example, a Microsoft Office® document), an electronic mail (email), downloaded web page, an instant messaging element in accordance with Session Initiation Protocol (SIP) or another messaging protocol, or the like.

According to one embodiment, the term “malware” may be construed broadly as any code or activity that initiates a malicious attack and/or operations associated with anomalous or unwanted behavior. For instance, malware may correspond to a type of malicious computer code that executes an exploit to take advantage of a vulnerability, for example, to harm or co-opt operation of a network device or misappropriate, modify or delete data. In the alternative, malware may correspond to an exploit, namely information (e.g., executable code, data, command(s), etc.) that attempts to take advantage of a vulnerability in software and/or an action by a person gaining unauthorized access to one or more areas of a network device to cause the network device to experience undesirable or anomalous behaviors. The undesirable or anomalous behaviors may include a communication-based anomaly or an execution-based anomaly, which, for example, could (1) alter the functionality of an network device executing application software in an atypical manner (a file is opened by a first process where the file is configured to be opened by a second process and not the first process); (2) alter the functionality of the network device executing that application software without any malicious intent; and/or (3) provide unwanted functionality which may be generally acceptable in another context. Additionally, malware may be code that initiates unwanted behavior which may be, as one example, uploading a contact list from an endpoint device to cloud storage without receiving permission from the user.

The term “shellcode” refers to a small piece of executable code that resides in data (e.g., is injected into data), is used as a payload of malware, and, in some cases, contains a shell command to execute an exploit.

In certain instances, the term “detected” is used herein to represent that there is a prescribed level of confidence (or probability) on the presence of an exploit or attack within an object under analysis.

The term “network device” should be construed as any electronic device with the capability of connecting to a network. Such a network may be a public network such as the Internet or a private network such as a wireless data telecommunication network, wide area network, a type of local area network (LAN), or a combination of networks. Examples of a network device may include, but are not limited or restricted to, a laptop, a mobile phone, a tablet, a computer, etc.

The term “computerized” generally represents that any corresponding operations are conducted by hardware in combination with software and/or firmware. Also, the terms “compare” or “comparison” generally mean determining if a match (e.g., a certain level of correlation) is achieved between two items where one of the items may include a particular signature pattern.

Finally, the terms “or” and “and/or” as used herein are to be interpreted as inclusive or meaning any one or any combination. Therefore, “A, B or C” or “A, B and/or C” mean “any of the following: A; B; C; A and B; A and C; B and C; A, B and C.” An exception to this definition will occur only when a combination of elements, functions, steps or acts are in some way inherently mutually exclusive.

As this invention is susceptible to embodiments of many different forms, it is intended that the present disclosure is to be considered as an example of the principles of the invention and not intended to limit the invention to the specific embodiments shown and described.

II. Early Warning Alert Generation Methodology

A. Operational Flow for Early Warning Alert Generation

Referring to FIG. 1, an exemplary logical representation of a cyber attack early warning system 100 (hereinafter “EWS”) is shown. Herein, some or all of various incoming objects associated with monitored network traffic are received by an exemplary malicious content detection (MCD) system 102, as described in U.S. patent application Ser. No. 14/042,483, filed Sep. 30, 2013 and incorporated herein by reference in its entirety, which preferably comprises virtual execution logic that is part of a dynamic analysis engine, either directly or via an optional static analysis engine. According to one embodiment of the disclosure, the static analysis engine is configured as a capture and filter device that receives the incoming objects and conducts heuristics, exploit signature checks and/or vulnerability signature checks on some or all of the objects, to identify “suspicious” objects having one or more characteristics associated with an exploit. It is envisioned that the MCD system 102 may also perform a behavioral analysis of the incoming objects. For example, the MCD system 102 may analyze one or more objects to determine how the object was received, e.g., via browser download, email attachment, external drive, SQL injection compromise, and the like, without limitation. Furthermore, certain artifacts that may be unique with respect to a type of malware attack may be analyzed. The behavioral analysis may also consider propagation mechanisms of an object, to determine how the object communicates or navigates across and/or through a network, for example.

When suspicious objects are identified by the MCD system 102, an attack alert 110 is generated and received by the EWS 100. The attack alert 110 may comprise analysis logs detailing certain characteristics of exploits, such as from one or more virtual machines, attack metadata, communication logs from the client device and/or the MCD system 102, and various other information corresponding to a malware attack, without limitation.

As shown in FIG. 1, the EWS 100 comprises an input engine 120, an attack-specific engine 130, a correlation engine 145, and a reporting engine 155. According to one embodiment of the disclosure, each of the input engine 120, the attack-specific engine 130, the correlation engine 145, and the reporting engine 155 may correspond to one or more software modules stored in non-transitory storage medium that, when executed by one or more hardware processors, performs the below identified functionality. Alternatively, some or all of these engine 120, 130, 145 and 155 may correspond to hardware logic. Independent of the system architecture, the communicative coupling between the input engine 120, attack-specific engine 130, correlation engine 145, and reporting engine 155 is provided in a generally serial configuration. However, it is contemplated that some or all of the aforementioned engines may be provided in a parallel configuration, where the attack alert 110 may be processed concurrently, or in any other configuration such as a daisy-chain or any other topology without limitation.

In one embodiment, the EWS 100 and MCD system 102 are communicatively coupled with a threat intelligence network 105. The threat intelligence network 105 may be configured to store and access data regarding malware attacks across a number of objects, including for example, web-based, email-based, and file-based threats. Moreover, it is envisioned that the threat intelligence network 105 may be configured to store historical information regarding previously analyzed and/or known malware attacks. The threat intelligence network 105 may also be periodically and/or aperiodically updated so as to store information regarding new malware attacks, reports, alerts, and/or corresponding features, for example.

Once generated, some or all of the data associated with the attack alert 110 is routed to the EWS 100 using pull coding, such that the initial request for the attack alert 110 originates from the EWS 100, and is then responded to by a server, such as via the threat intelligence network 105. In one embodiment, using pull coding, the threat intelligence network 105 may be idle until the EWS 100 interrupts the threat intelligence network 105 with a request. Upon receipt of the request, the threat intelligence network 105 may become an active component. In one embodiment, the threat intelligence network 105 may be configured so as to determine its own interest in the request and also to consider the origin of the request prior to communicating any further data. It should be appreciated, however, that other routing methodologies such as push technologies, or some hybrid of push/pull technologies are within the scope and spirit of the present disclosure.

Upon receipt by the EWS 100, the attack alert 110 is communicated to the input engine 120 for analysis, feature extraction and classification using various logic, such as heuristics, exploit/vulnerability logic and parsing logic, for example, as explained in detail further below. When the attack alert 110 is classified, the input engine 120 generates a classified result 125 that is communicated to the attack-specific engine 130, which provides attack-specific analysis with respect to the classified result 125. As shown, the attack-specific engine 130 includes attack-specific logic 135 that may be specifically tailored to analyze one of various malware attacks, including by way of non-limiting example, APT, POS, and Crimeware attacks.

For example, the attack-specific logic 135 may comprise a plurality of analysis mechanisms such as one or more heuristic, probabilistic, and/or machine learning algorithms specific to a type of malware attack. Consequently, a particular attack alert may be compared with previously known attacks, incidences, feature sets, vulnerabilities, attack signatures, and the like, to determine correspondence with respect to a plurality of time-dependent and time-independent features. In one embodiment, the time dependent features may comprise the number of request-response sessions, and/or the time, day, and month of the attack. Similarly, in one embodiment, time-independent features may comprise the time-independent features such as the geographic location of the attack target, industry, employees of an organization, either alone or in combination, without limitation. The combination of time-dependent and time-independent features are then examined using a plurality of correlation techniques, so as to determine a degree of similarity based on the presence and/or volume of the time-dependent and time-independent features and generate an attack-specific engine result 140 that is communicated to the correlation engine 145.

Using probabilistic or machine-learning algorithms, the correlation engine 145 is configured to determine an “attack value”, which in one embodiment comprises a real-time value that may be based on at least the attack-specific engine result 140, time-dependent and time-independent features, and a consideration of potential attack targets, for example. The correlation engine 145 is configured to compare the attack value with a predetermined threshold value to determine whether an early warning alert should be generated.

In one embodiment, the predetermined threshold value may represent, for example, certain incidences, feature sets, vulnerabilities, and/or attack signatures specific to any of various malware attacks. More specifically, it is envisioned that the predetermined threshold value may comprise any plurality of characteristics of a particular malware attack, including by way of non-limiting example, the number and/or existence of various API calls, request-response sessions, and geographical information regarding the most recent, and/or previously targeted parties. It should be understood that the predetermined threshold value might vary for different types of malware attacks. Thus, for certain malware attacks, only one characteristic may be required to be present so that the attack value matches or exceeds the threshold value. Of course, in other instances, the threshold value may be adapted to accommodate any number of characteristics depending on the type of malware attack.

In general, if the attack value matches or exceeds the predetermined threshold value, then a correlated result 150 is generated and communicated to the reporting engine 155 for final processing and reporting. In the event that the attack value fails to exceed the predetermined threshold, then the correlated results 150 may indicate, for example, that the reporting engine 155 should not generate an early warning alert or report 160. The correlated results 150 may be stored in a database for future reference.

If the correlated results 150 indicate that an early warning alert or report 160 should be generated, then at least portions of the correlated results 150 and attack-specific engine results 140 may be combined and communicated to the reporting engine 155. The reporting engine 155 may issue an early warning alert or report 160 (e.g., an email message, text message, display screen image, etc.) to security administrators for example, communicating the urgency in handling and preferably preventing one or more predicted attacks.

Alternatively, the early warning alert or report 160 may trigger further analysis of the object by the MCD system 102, for example, to verify behavior of the object as an exploit. The early warning alert or report 160 may also include detailed instructions pertaining to specific attack types, potential issues thereto, security holes, and best practices to prevent one or more predicted malware attacks. It should be appreciated that the reporting engine 155 may also be configured to update the threat intelligence network 105 with information corresponding to the instantly analyzed attack alert for future reference and/or further processing.

B. General Architecture of Network Device Deploying Early Warning Alert Generation Logic

Referring to FIG. 2A, an exemplary block diagram of an EWS detection environment 200 communicatively coupled to a communication network 261 is shown. The EWS detection environment 200 comprises a server device 260, an optional firewall 262, an optional network interface 270, a MCD system 263, a client device 264, and a threat intelligence network 265 that is communicatively coupled with the MCD system 263 and the exemplary EWS 100.

According to the embodiment illustrated in FIG. 2A, the MCD system 263 is a network device that is adapted to analyze information associated with network traffic routed over a communication network 261 between at least one server device 260 and at least one client device 264. The communication network 261 may include a public network such as the Internet, in which case an optional firewall 262 (represented by dashed lines) may be interposed on the communication path between the public network and the client device 264. Alternatively, the communication network 261 may be a private network such as a wireless data telecommunication network, wide area network, a type of local area network (LAN), or a combination of networks.

As shown, the EWS 100 may be communicatively coupled with the communication network 261 via a network interface 270. In general, the network interface 270 operates as a data capturing device (sometimes referred to as a “tap” or “network tap”) that is configured to receive data propagating to/from the client device 264 and provide at least some of this data to the MCD system 263. Alternatively, the MCD system 263 may be positioned behind the firewall 262 and in-line with client device 264.

According to one embodiment of the disclosure, the network interface 270 is capable of receiving and routing objects associated with network traffic to the MCD system 263. The network interface 270 may provide the entire traffic or a certain subset of the network traffic, for example, such as one or more files that are part of a set of flows, packet payloads, or the like. In some embodiments, although not shown, network interface 270 may be contained within the MCD system 263.

As further shown in FIG. 2A, the EWS 100 comprises input engine 210, attack-specific engine 220, correlation engine 230, and reporting logic 240. Although the components disposed within the EWS 100 are shown in a communicatively coupled serial configuration, it is envisioned that other topologies may also be implemented, such as, for example, parallel and daisy-chain configurations. It should be appreciated that the input engine 210, attack-specific engine 220, correlation engine 230, and reporting logic 240 may each be separate and distinct components, but the combination of components may also be implemented in a single block and/or core.

In some embodiments, as best shown in FIG. 2B, the input engine 210 comprises one or more software modules that, when executed by one or more processors, receives the attack alert 110 from the threat intelligence network 265 for classification, prior to being communicated to the attack-specific engine 220. In some embodiments, as shown in FIG. 2A, the input engine 210 may comprise one or more software modules such as attack alert logic 211, when executed by one or more processors, extracts certain data from the attack alert 110 so as to classify the attack using heuristics, exploit signature checks and/or vulnerability signature checks, for example. It should be understood that the attack alert logic 211 and the input engine may be one or more software modules executed by the same processor or different processors, where these different processors may be located within the same processor package (e.g., different processor cores) and/or located at remote or even geographically remote locations that are communicatively coupled (e.g., by a dedicated communication link) or a network.

Referring again to FIG. 2A, the attack alert logic 211 may comprise modules dedicated to certain tasks, such as, by way of non-limiting example, heuristics logic 212, exploit/vulnerability matching logic 214, and/or parsing logic 215. The heuristics logic 212 may be adapted for analysis of certain portions of network traffic under analysis to determine whether any portion thereof corresponds to either (i) a “suspicious” identifier such as either a particular Uniform Resource Locator “URL” that has previously been determined as being associated with known exploits, a particular source or destination (IP or MAC) address that has previously been determined as being associated with known exploits; (ii) a particular exploit pattern; or (iii) a particular shellcode pattern, alone or in combination, without limitation. When deployed, the attack alert logic 211 may be adapted to perform exploit signature checks, which may involve a comparison of an object under analysis against one or more pre-stored exploit signatures (e.g., pre-configured and predetermined attack patterns) from database 216. Additionally or in the alternative, the attack alert logic 211 may be configured with parsing logic 215 so as to separate various aspects of the attack alert 110 for more efficient and timely processing. For example, the parsing logic 215 may be configured so as to receive data associated with the attack alert 110 to construct a data structure, including by way of non-limiting example, a parse tree, abstract syntax tree or any other hierarchical structure, thereby providing a structural representation of data associated with the attack alert 110.

Additionally or in the alternative, the attack alert logic 211 may be configured with exploit/vulnerability matching logic 214 that is adapted to perform vulnerability signature checks, namely a process of uncovering deviations in messaging practices set forth in applicable communication protocols (e.g., HTTP, TCP, etc.). The term “signature” designates an indicator of a set of characteristics and/or behaviors exhibited by one or more exploits that may not be unique to those exploit(s). Thus, a match of the signature may indicate to some level of probability, often well less than 100%, that the attack alert 110 comprises a certain exploit. In some contexts, those of skill in the art have used the term “signature” as a unique identifier or “fingerprint,” for example, of a specific virus or virus family (or other exploit), which is generated for instance as a hash of its machine code, and that is a special sub-case for purposes of this disclosure.

As further shown in FIG. 2A, once the attack alert 110 has been analyzed, the input engine 210 generates classified results 125 based on determinations of the attack alert logic 110. The classified results 125 may, among others, classify any malware and/or exploits detected into a family of malware and/or exploits, describe the malware and/or exploits and provide the metadata associated with any object(s) within which the malware and/or exploits were detected. Furthermore, the attack alert logic 211 may store the classified results 125 in the database 216 for future reference.

As shown, the classified results 125 are communicated to the attack-specific engine 220. The attack-specific engine 220 comprises attack-specific logic 135 that comprises a plurality of attack cores 222₁-222_N, each specific to one of various types of malware attacks, including by way of non-limiting example, APT, POS, and Crimeware attacks. The attack-specific engine 220 further comprises routing logic 223 that is configured so as to route the classified result 125 to one of the appropriate attack cores 222₁-222_N. It is envisioned that each of the attack cores 222₁-222_N is configured so as to analyze the content of the attack alert 110 and apply a plurality of analysis mechanisms specific to the type of malware attack. In one embodiment, each of the attack cores 222₁-222_N may be configured as a plug-in, extension, and/or an add-on, for example, so as to apply analysis mechanisms specific to the type of malware attack. Each of the plug-ins may include feature logic 224 specific to each type of attack, including, but not limited to network and operating system behavior when a client device and/or network is under a specific type of malware attack, as explained in further detail below. It is envisioned that each of the plug-ins may be configured so as to cooperate with any real-time data collection and/or processing capabilities of the EWS 100.

In one embodiment, each of the plurality of attack cores 222₁-222_N comprises feature logic 224 specific to each type of attack, based on, for example, attacks stored in a feature-set database 225 comprising historical data corresponding to known and/or previously analyzed attacks. More specifically, the feature logic 224 may comprise time-dependent features and time-independent features. With respect to time-dependent features, a particular attack alert may be compared with previously known attacks, incidences, attack signatures, and the like, so as to determine correspondence with respect to a plurality of time-dependent features such as the number of methods for a request-response over a predetermined period of time, the time of day of the attack, the duration of the attack, the month, year, execution time of the attack, and the like, without limitation. Furthermore, a plurality of time-independent features, as the name implies, may comprise features such as the geographic location of the attack target, industry, role/title of a specific party etc., without limitation. For example, in one embodiment, in the event that there is an APT attack, the feature-set database 225 may have information stored thereon relating to known APT attacks, such as the medium most often used in such attacks, e.g., a document, or PDF file, and the like; along with any correspondingly suspicious instructions, commands, and/or processes.

As shown in FIG. 2A, each of the attack cores 222₁-222_N further comprises one or more logic models 228 for further analysis of the classified result 125. It should be appreciated that the logic models 228 may comprise one or more software modules such as probabilistic logic, heuristic logic, and deterministic logic, that, when executed by one or more processors, extracts certain data from the classified result 125 so as to analyze it, at least in part, based on attack-specific (i) pattern matches; (ii) heuristic, probabilistic, or determinative analysis results; (iii) analyzed deviations in messaging practices set forth in applicable communication protocols (e.g., HTTP, TCP, etc.); (iv) analyzed compliance with certain message formats established for the protocol (e.g., out-of-order commands); and/or (v) analyzed header or payload parameters to determine compliance.

It should be understood that the logic models 228 may comprise one or more software modules executed by the same processor or different processors, where these different processors may be located within the same processor package (e.g., different processor cores) and/or located at remote or even geographically remote locations that are communicatively coupled (e.g., by a dedicated communication link) or a network. Once the attack-specific engine 220 has completed its analysis, an attack-specific result 140 comprising certain attack features is generated and communicated to the correlation engine 230. Furthermore, the attack-specific logic 135 may store the attack-specific result 140 in the feature-set database 225 for future reference.

The correlation engine 230 comprises alert value determination logic 231 and a database 232. The alert value determination logic 231 is communicatively coupled to the database 232, where it is envisioned that information regarding potential targets may be stored. Notable information that may be stored on the database 232 includes at least real time features corresponding to potential targets, such as geographic location, industries, segments, whether the target stores credit card information or other sensitive personal data, previous known attacks, alerts, current events, and specifics regarding the various targets' officers, board of directors, and the like.

The alert value determination logic 231 of the correlation engine 230 is configured using at least one of a probabilistic or machine-learning algorithm to determine, in real time, an attack value that represents a probability (or level of confidence) that a potential target may also be attacked. In doing so, the correlation engine 230 is configured so as to consider the attack features learned from the attack-specific engine 220, and real time features stored on the database 232. It is envisioned that the alert value determination logic 231 shall determine if the attack value matches or exceeds a predetermined threshold, so as to generate a correlated result 150 that indicates whether or not a probable target has been found, and therefore alerted with respect to the potential of a malware attack.

Consequently, the correlated result 150 is communicated to the alert generation logic 242 of the reporting logic 240, which may generate an alert for the client device 264 and/or route the alert to the threat intelligence network 265 for further analysis. In addition, the alert may be routed to the communication network 261 for further analysis by a network administrator, for example. The reporting logic 240 may issue an early warning alert or report 160 (e.g., an email message, text message, display screen image, etc.) to security administrators for example, communicating the urgency in handling one or more predicted attacks. The early warning alert or report 160 may trigger a further analysis of the object to verify the behavior of the object as an exploit. It is envisioned that the early warning alert or report 160 may also comprise instructions so as to prevent one or more predicted malware attacks. Finally, the reporting logic 240 may store the correlated results 150 in the database 242 for future reference.

Referring now to FIG. 2B, an exemplary embodiment of a logical representation of the cyber attack early warning system 100 of FIG. 2A is shown. In one embodiment, a network appliance 285 comprises a housing 280, which is made entirely or partially of a rigid material (e.g., hardened plastic, metal, glass, composite or any combination thereof) that protect circuitry within the housing 280, namely one or more processors 275 that are coupled to communication interface logic 278 via a first transmission medium 284. Communication interface logic 278 enables communications with other cyber attack early warning systems 100 and/or the threat intelligence network 265 of FIG. 2A, for example. According to one embodiment of the disclosure, communication interface logic 278 may be implemented as a physical interface including one or more ports for wired connectors. Additionally, or in the alternative, communication interface logic 278 may be implemented with one or more radio units for supporting wireless communications with other electronic devices.

Processor(s) 275 may further be coupled to persistent storage 290 via a second transmission medium 286. According to one embodiment of the disclosure, persistent storage 290 may include the cyber attack early warning system 100, which in one embodiment comprises (a) input engine 210; (b) attack-specific engine 220; (c) correlation engine 230; and reporting logic 240. It is envisioned that one or more of these engines (or logic units) could be implemented externally from the cyber attack early warning system 100 without extending beyond the spirit and scope of the present disclosure.

Referring now to FIG. 3, a flowchart of an exemplary method for providing early warning of a potential attack on a system or device based on a real-time analysis. In block 301, upon detection of potentially malicious network traffic, the MCD system is configured to generate an attack alert. In block 302, the attack alert is received by the EWS 100 and communicated to the input engine 120. In block 303, the input engine 120 performs feature extraction and classification with respect to the attack alert. For example, the attack alert may correspond to any of various malware attack types, including without limitation APT, POS, Crimeware attacks, and the like. In block 304, once the attack alert is classified, the input engine's result is communicated to the attack-specific engine, which provides further in-depth analysis of the attack alert. As noted herein, the expert engine comprises a plurality of attack cores, each specific to one of various malware attacks, including by way of non-limiting example, APT, POS, and Crimeware attacks.

In block 305, each of the attack cores is configured to analyze the content of the attack alert and apply a plurality of analysis mechanisms specific to the type of malware attack. For example, utilizing the attack cores, a particular attack alert may be compared with previously known attacks, incidences, feature sets, attack signatures, etc. to determine correspondence with respect to a plurality of time-sensitive features such as the number of methods for a request-response, and/or the time of day of the attack. A plurality of time-insensitive features such as geographic location of the present attack target, industry, etc. may also be analyzed, without limitation.

In block 306, after analyzing the combination of time-sensitive and time-insensitive features, an attack-specific result is generated and communicated to the correlation engine so as to correlate the result using real-time features across a plurality of potential targets. In block 307, using probabilistic and/or machine-learning algorithms, the correlation engine is configured to determine an “attack value” based on the attack-specific result, and a consideration of probable attack targets. For example, probable attack targets may include companies in the same industry, geographic region, and/or companies having the same size/scale of a presently attacked target, without limitation. In block 308, if the determined attack value matches or exceeds a predetermined threshold value, then an early warning alert is provided in block 310 and communicated to probable targets of a malware attack. If the attack value is less than the predetermined threshold, then, at block 309, the early warning alert is not generated. In either event, however, the threat intelligence network 265 is updated with respect to any potential incident, and the corresponding results thereto.

C. Exemplary Alert

Referring to FIG. 4, a network device may be configured to receive an early warning alert or report 160. In FIG. 4, for illustrative purposes, the network device is represented as a mobile network device 400 (e.g., smartphone, tablet, laptop computer, netbook, etc.). The mobile network device 400 includes a display screen 410; a receiver and/or transmitter (e.g. transceiver) such as an antenna 411.

In one embodiment, the exemplary alert 450 (e.g., an email message, text message, display screen image, etc.) is communicated to security administrators for receipt/viewing on the mobile network device 400. For example, the exemplary alert 450 may indicate the urgency in handling one or more predicted attacks. Furthermore, the exemplary alert 450 may comprise instructions so as to prevent one or more predicted malware attacks. The exemplary alert 450 may also comprise information with respect to the origination of the potential attack, along with suspicious behavior that might confirm the attack with respect to a potential target.

In the foregoing description, the invention is described with reference to specific exemplary embodiments thereof. However, it will be evident that various modifications and changes may be made thereto without departing from the broader spirit and scope of the invention as set forth in the appended claims.

Claims

1. A computerized method, comprising: receiving, by a first networked system, an attack alert associated with an object analyzed and identified as suspicious; andextracting, by the first networked system, features from the attack alert;classifying, based on the extracted features, the attack alert by identifying the suspicious object as being at least a type of malware attack;providing, by the first networked system, data associated with the classified attack alert to an attack-specific engine that conducts a plurality of analyses on the data and computes results for each of the plurality of analyses, wherein the attack-specific engine comprises a plurality of attack cores and each attack core of the plurality of attack cores being configured to (i) conduct at least an analysis of the plurality of analyses on a portion of the data associated with the classified attack alert in accordance with at least a corresponding type of malware attack and (ii) compute an attack-specific result of a plurality of attack-specific results, the attack-specific result including at least one or more features of the extracted features that pertain to the corresponding type of malware attack being analyzed by the attack core;receiving the attack-specific result and computing an attack value based on an analysis of the extracted features including time-dependent features and time-independent features corresponding to a potential target; andgenerating, by the first networked system, an alert directed to one or more predicted attacks on the potentional target based on the attack value.

2. The computerized method of claim 1, wherein the attack value represents a probability of the potential target being attacked.

3. The computerized method of claim 1, wherein the plurality of attack cores corresponds to a first attack core, the first attack core being configured to (i) extract the features from the attack alert and (ii) compute the attack-specific result including one or more features of the extracted features that pertain to a first type of malware attack being analyzed by the first attack core.

4. The computerized method of claim 1, wherein the potential target is associated with an industry type or geographic location.

5. The computerized method of claim 1, wherein the extracted features comprise at least one of various analysis logs from at least one of (i) one or more virtual machines, (ii) the metadata associated with the malware attack, or (iii) communication logs from a malicious content detection system.

6. The computerized method of claim 1, wherein the time-dependent features comprise at least one of the number of request-response sessions over a predetermined period of time, a time of day of the malware attack, a duration of the malware attack, or an execution time of the malware attack.

7. The computerized method of claim 1, wherein the time-independent features comprise at least one of a geographic location of the potential target, an industry, or a role or title of a specific party.

8. The computerized method of claim 1, wherein the attack value is generated based on the attack-specific result from each of the plurality of attack cores.

9. The computerized method of claim 1, wherein the alert being generated based on a comparison of the attack value with a predetermined threshold value to determine whether the alert is to be generated.

10. The computerized method of claim 1, wherein the first networked system corresponds to an early warning system.

11. A system comprising: one or more processors; anda storage module communicatively coupled to the one or more processors, the storage module including an input engine to (i) receive data associated with an attack alert associated with an object analyzed and identified as suspicious by a remote source, the attack alert including information associated with the suspicious object having one or more characteristics associated with a malware attack, and (ii) classify the received data to produce a classified result, the classified result is to identify at least a type of a malware attack,an attack-specific engine communicatively coupled to the input engine, wherein the attack-specific engine comprises a plurality of attack cores each directed to analyze the classified result according to at least a specific type of malware attack and at least a first attack core of the plurality of attack cores being configured to compute an attack-specific result including one or more features that pertain to the identified malware attack being monitored by the first attack core, anda correlation engine communicatively coupled to the attack-specific engine, the correlation engine to compute an attack value based on at least the attack-specific result and a consideration of potential attack targets, wherein the attack value represents a probability of a potential target being attacked and is compared to a threshold value to determine whether or not to generate an alert.

12. The system of claim 11, wherein the attack-specific engine comprises the plurality of attack cores and routing logic configured to route the classified result to at least one of the plurality of attack cores, each of the plurality of attack cores being configured as a plug-in that compares the one or more features included in the attack-specific results with features associated with a known malware attack.

13. The system of claim 11, wherein the attack-specific engine is configured to analyze the one or more features included in the attack-specific results to one or more of time-dependent features and one or more time-independent features.

14. The system of claim 13, wherein the one or more time-dependent features include at least one of (i) a time of day of the identified malware attack or (ii) a duration of the identified malware attack while the time-independent features include at least one of (i) a geographic location of a target of the identified malware attack or (ii) an industry attacked by the identified malware attack.

15. The system of claim 11, wherein the alert comprises at least one of an email message, text message, or a display screen image.

16. The system of claim 11, wherein the alert is communicated to one or more security administrators to indicate the urgency in handling one or more predicted malware attacks.

17. The system of claim 13, wherein the one or more time-dependent features comprise at least one of the number of request-response sessions over a predetermined period of time, a time of day of the malware attack, a duration of the malware attack, or the execution time of the malware attack; andwherein the one or more time-independent features comprise at least one of a geographic location of the malware attack target, an industry, or a role or title of a specific party.

18. The system of claim 11, wherein each attack core of the plurality of attack cores comprises feature logic specific to analyze a type of malware attack.

19. The system of claim 18, wherein the first attack core of the plurality of attack cores to analyze a different type of malware attack than a second attack core of the plurality of attack cores.

20. The system of claim 11, wherein each attack core of the plurality of attack cores comprises one or more logic models that, when executed, extracts certain data from the classified result for analysis to generate the attack-specific result.

21. The system of claim 11, wherein the first attack core of the plurality of attack cores being configured to analyze the information associated with the attack alert to information associated with previously known attacks, incidences or feature sets.

22. A computerized method, comprising: receiving, by a first networked system, an attack alert associated with an object analyzed and identified as suspicious, the attack alert including information associated with the suspicious object having one or more characteristics associated with an exploit or other attack; andproviding the information associated with the exploit or other attack to an attack-specific engine that conducts a plurality of analyses directed to one or more types of attacks and computes results for each of the plurality of analyses each being provided as an attack-specific result;receiving the attack-specific results and computing an attack value based on an analysis of extracted features from the attack alert including time-dependent features and time-independent features with features corresponding to potential targets; andgenerating, by the first networked system, an alert directed to one or more predicted attacks based on the attack value,wherein the attack-specific engine comprises a plurality of attack cores, each of the plurality of attack cores being directed to an analysis for at least a specific type of attack and at least a first attack core of the plurality of attack cores being configured to (i) extract features from the attack alert and (ii) compute an attack-specific result including one or more features of the extracted features that pertain to a first type of attack being monitored by the first attack core.

23. The computerized method of claim 22, wherein the attack value represents a probability of a potential target being attacked.

24. The computerized method of claim 22, wherein the one or more types of attacks corresponding to malware attacks.

25. The computerized method of claim 22, wherein the first type of attack includes a particular type of malware attack.