Methods and cloud-based systems for detecting malwares by servers
Disclosed herein are systems and method for detecting malwares by a server of a sandbox. In one aspect, an exemplary method comprises receiving, by a deep dynamic analysis tool of the server, a sample of a process from an endpoint device with a request for a final verdict indicative of whether the process is a malware or clean based on a deep dynamic analysis, collecting events for the sample, the collected events including events collected using at least one invasive technique, analyzing the collected events using one or more detection models of the deep dynamic analysis tool to detect malwares and issue the final verdict, and sending final verdict to the endpoint device from which the sample is received.
1. A method for detecting malwares by a server of a sandbox, comprising:
receiving, by a deep dynamic analysis tool of the server, a sample of a process from an endpoint device with (1) a request for a final verdict indicative of whether the process is a malware or clean based on a deep dynamic analysis and (2) Type A events collected without invasive techniques by the endpoint device;
collecting, by the deep dynamic analysis tool of the server, Type B events for the sample, the Type B events are collected using at least one invasive technique by the server;
analyzing, by the deep analysis tool of the server, the Type B events using one or more detection models of the deep dynamic analysis tool to detect malwares and issue the final verdict;
sending the Type A events, the Type B events, and the final verdict to a correlator configured to:
correlate the Type B events with the Type A events;
for events of Type A that are correlated with events of Type B used to detect malwares, create suspicious patterns of events of Type A;
sending, by the deep analysis tool of the server, the final verdict and the suspicious patterns to the endpoint device from which the sample is received.
2. The method of claim 1 , wherein the Type B events include at least one of: events based on virtualization technologies, and events based on modification of system components.
3. The method of claim 1 , wherein the Type A events are collected by an external monitoring tool.
4. The method of claim 1 , wherein the Type A events include at least one of: guaranteed events, logging events, tracing events, and noninvasive monitoring events.
5. The method of claim 1 , further comprising:
updating malware detection models of one or more endpoint devices and one or more servers of the sandbox based on results of the correlation, wherein the updating includes at least distributing the created suspicious patterns to the one or more endpoint devices and the one or more servers of the sandbox, the endpoint device from which the sample is received being one of the one or more endpoint devices to which the created suspicious patterns are distributed.
6. A system for detecting malwares, comprising:
at least one processor of a server of a sandbox configured to:
receive, by a deep dynamic analysis tool of the server, a sample of a process from an endpoint device with (1) a request for a final verdict indicative of whether the process is a malware or clean based on a deep dynamic analysis and (2) Type A events collected without invasive techniques by the endpoint device;
collect, by the deep dynamic analysis tool of the server, Type B events for the sample, the Type B events are collected using at least one invasive technique by the server;
analyze, by the deep analysis tool of the server, the Type B events using one or more detection models of the deep dynamic analysis tool to detect malwares and issue the final verdict;
send the Type A events, the Type B events, and the final verdict to a correlator configured to:
correlate the Type B events with the Type A events;
for events of Type A that are correlated with events of Type B used to detect malwares, create suspicious patterns of events of Type A;
send, by the deep analysis tool of the server, the final verdict and the suspicious patterns to the endpoint device from which the sample is received.
7. The system of claim 6 , wherein the Type B events include at least one of: events based on virtualization technologies, and events based on modification of system components.
8. The system of claim 6 , wherein the Type A events are collected by an external monitoring tool.
9. The system of claim 6 , wherein the Type A events include at least one of: guaranteed events, logging events, tracing events, and noninvasive monitoring events.
10. The system of claim 6 ,
wherein the correlator is further configured to:
update malware detection models of one or more endpoint devices and one or more servers of the sandbox based on results of the correlation, wherein the updating includes at least distributing the created suspicious patterns to the one or more endpoint devices and the one or more servers of the sandbox, the endpoint device from which the sample is received being one of the one or more endpoint devices to which the created suspicious patterns are distributed.
11. A non-transitory computer readable medium storing thereon computer executable instructions for detecting malwares by a server of a sandbox, including instructions for:
receiving, by a deep dynamic analysis tool of the server, a sample of a process from an endpoint device with (1) a request for a final verdict indicative of whether the process is a malware or clean based on a deep dynamic analysis and (2) Type A events collected without invasive techniques by the endpoint device;
collecting, by the deep dynamic analysis tool of the server, Type B events for the sample, the Type B events are collected using at least one invasive technique by the server;
analyzing, by the deep analysis tool of the server, the Type B events using one or more detection models of the deep dynamic analysis tool to detect malwares and issue the final verdict;
sending the Type A events, the Type B events, and the final verdict to a correlator configured to:
correlate the Type B events with the Type A events;
for events of Type A that are correlated with events of Type B used to detect malwares, create suspicious patterns of events of Type A;
sending, by the deep analysis tool of the server, the final verdict and the suspicious patterns to the endpoint device from which the sample is received.
12. The non-transitory computer readable medium of claim 11 , wherein the Type B events include at least one of: events based on virtualization technologies, and events based on modification of system components.
13. The non-transitory computer readable medium of claim 11 , wherein the Type A events include at least one of: guaranteed events, logging events, tracing events, and noninvasive monitoring events.
14. The non-transitory computer readable medium of claim 11 , the instructions for detecting malwares by the server of the sandbox, including further instructions for:
updating malware detection models of one or more endpoint devices and one or more servers of the sandbox based on results of the correlation, wherein the updating includes at least distributing the created suspicious patterns to the one or more endpoint devices and the one or more servers of the sandbox, the endpoint device from which the sample is received being one of the one or more endpoint devices to which the created suspicious patterns are distributed.