FOG COMPUTING

DONE BY,

H.GAAYATHRIANAND INSTITUTE OF HIGHER TECHNOLOGYB.TECH IT-III YEAR

CONTENTS:• WHAT IS FOG COMPUTING?• WHAT IS THE NEED FOR FOG COMPUTING?• CLOUD vs FOG• APPLICATIONS• FOG COMPUTING AND THE INTERNET of THINGS• SECURITY AND PRIVACY ISSUES • CONCLUSION

• Transmitting and processing data requires bandwidth, The more data, the more bandwidth we need.• In cloud, Data transmission simply isn’t fast enough.•Reduces Latency & Improves QOS

WHY DO WE NEED FOG COMPUTING?

LIMITATIONS OVERCAME BY FOG:

• Fogging basically helps cloud systems by easing the burden of processing data.• Fogging provides an improved security of

encrypted data as it stays closer to the end user.•Conserving the Network Bandwidth by reducing the amount of data that is analysed and stored in a cloud.

FOG COMPUTING KEEPS THE DATA RIGHT WHERE THE INTERNET OF THINGS NEEDS IT

FOG COMPUTING FOR IoT

• The concept of “fog computing” has been introduced as a bridge between IoT devices in the field and remote data centers.• IoT devices can produce huge data sets that need to be

processed. With fog computing, some of that processing load can be handled by computing resources at the edge, by filtering and summarizing the data to reduce volume and increase value and relevance.• This shift from cloud to fog makes complete sense.• In doing so,it eliminates the bandwidth bottleneck and

cripples the iot in the long run.

SECURITY AND PRIVACY ISSUES

Man-in-the-middle Attack

In fog computing, there is the potential of a typical man-inthe-middle. In this attack, gateways serving as fog devices may be compromised or replaced by fake ones. In some scenarios, it is difficult to protect communication between fog node and IoT devices using encryption method. Encryption and decryption methods consume large amount of battery on mobile device.

Malicious Detection Technique in Fog Computing EnvironmentsWhen some fog nodes are compromised, hybrid detection technique is useful to detect malicious code in fog nodes. It is combined with signature-based detection technique [4] and behavior-based detection technique [5]. However, IoT devices have low computing power. The behavior-based detection technique costs largely overhead. While signature-based detection technique is more efficient than it. However, it is difficult to detect various forms of the malicious code in fog nodes.

Malicious Fog Node Problem

In order to provide service to user, fog nodes process data received from the IoT devices. If the workloads is heavy, it is divided into several jobs and processed by several fog nodes. If some fog nodes are compounded by a malicious user, it is difficult to ensure the integrity of the data.

Data Protection

Messages generated from IoT devices is sent to the nearest fog nodes. It is difficult to process a volume of data on IoT devices. The data is divided into some parts and sent to several fog nodes to process it. At this point, the contents of the data should be analyzed without exposing it. When distributed and processed data is merged, the integrity of the data should be guaranteed. Because of limited resources, it is difficult to encrypt or decrypt data on IoT device.

SOLUTIONS

MAN-IN-THE-MIDDLE Attack• Traditional anomaly detection methods can expose man-in-the-middle attack Malicious Detection Technique in Fog Computing Environments• In order to complement this problem, some computation of behavior-based detection techniques

that is running on the cloud is distributed into fog nodes.• Suspected malware files on a fog node are sent to cloud.• If the file is new malware, a result of analysis is stored in the signature database. • The result is transmitted to the fog node and updated signature information. Malicious Fog Node Problem• Before the computation begin, fog nodes must be trusted each other.• An authentication protocol is required. • Fog nodes should trust cloud, because there is no a fog node that manages other fog node. • In order to process a volume of data, fog nodes which is authenticated by the cloud should be only

located in fog environment.Data Protection• light-weight encryption algorithms or masking techniques are required.

CONCLUSION• Fog computing is more resilient, rugged and secure • However the high end business computing would require

large amounts of resources while not being very delay sensitive. These kinds of jobs can be handled using traditional cloud computing systems successfully more than fog nodes. • Hence fog computing will never be able to replace cloud

computing and become the sole cloud computing model of the future. • Thus, it is safe to state that cloud and fog would exist side by side serving two different communities and complement each other where necessary.

THANK YOU