This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
As cloud computing provides many benefits to the organizations, shift towards the cloud is increased more now days. Security is becoming more concerned factor for cloud environment. A key approach to secure cloud computing is that data owner stores the data in the encrypted format in the cloud and provide decryption technique to the user of that data. Internet technology is growing rapidly day by day. With the increase of this technology the amount of data transmission is also increased in huge amount. More and more communication is based upon the multimedia data i.e. video or images. Encryption of video data becomes more challenging because encrypting all compressed video has some drawbacks or more time consuming, so many selective encryption algorithms came into focus. In this paper we discussed the novel selective video encryption algorithm that is providing more security level with encrypting fewer amounts of data. Also we discuss how it can be implemented in the cloud environment to provide security to video data stored over the cloud database.
Keywords: video compression, cloud databse.
Secure storage of data is need of every organization, businesses, banking sector etc. Every organization has some private and confidential data that should not seen by some other party. So organization needs to protect their data by using any technique. Many algorithms are proposed for the video encryption so far.
Data outsourcing is the most growing field that allows user and organizations to give their data to external vendors who are responsible for storage and managing that data. This helps the organizations to concentrate only on their core business rather than management and storage of data. This also reduces their cost required for maintaining hardware, storing the data and maintaining that. Though outsourcing has many benefits it also introduces many concerns about security. This brings many challenges in data security as the organizations put their private data on the many shared servers which are not under the control of data owner. Third party i.e. the provider of that storage has the overall responsibility of the storing, managing and securing the data.It is proposed that limiting the access of information is not the proper way to secure the data. This proposal gives the concept of ââ‚¬ËœInformation accountabilityââ‚¬â„¢ . Information accountability has more advantage than the concept of restricting the information from the user. Information accountability successfully used from many years .
Health Insurance Portability and Accountability Act-HIPAA  is relevant for privacy and security of health data and U.S. Public Law known as Sarbanesââ‚¬"Oxley Act  is relevant to certify security and accuracy of financial data.
Review of literature
Many algorithms have been proposed till now for video encryption. In this section we discussed some of the techniques that are proposed by different authors.
As proposed by Hatem M. Abdul Kader, AES is used for encryption which compares the execution of the AES with respect to two other methods over a point-to-point channel. This paper showed AES encryption technique can be used more effectively for MPEG-4. The execution of this AES encrypting frames is quite enough to show the received frames on time. This encrypting method delays overhead is less than the of XOR or RC4 technique.
Another technique i.e. Light weight Encryption technique proposed by Dr. Florence. It mainly focuses on accomplishing high security over data at a low computing time. . it was carried out by encrypting Intra frames using DCT and DWT . It abides of avoiding the computationally choosing compensation step and tends to exploit the redundancy in the frames of a video. By converting group of pictures to a single picture with high spatial correlation and after conversion, all these inter frames are then scattered which reduces the computing time. There is a technique (using LFSR and the GF polynomial), the size using key space can be increased, that brings out robustness.
K. Bhagalaxmi proposed an algorithm for video encryption. This paper suggests about the selectively encrypting a fraction of whole video. The author believes this can be widely used in pay per view program or any online video on demand applications.
Normally video is processed in a compressed format. Many techniques are used for video compression. Here, in this paper we are considering the MPEG-2 format. Before we go for encryption part we should understand the MPEG-2 format. In this section we will discuss about the MPEG-2 format and then discuss about different parameters that are to be considered for the encryption.
MPEG-2 developed by ISO/IEC/JTC/SC29/WG11. It supports HDTV quality and real time transmission. MPEG-2 video is consists of four layers such as: Group of Pictures, slice, Macroblock, block. GOP consists of the header and sequence of frames or pictures. Pictures are known as the primary unit of videos. The sequence of pictures forms the video. Macroblock is the basic unit of MPEG format. It consists of 16X16 pixel segment. One Macroblock consists of 4 blocks. Block is a smallest unit of MPEG a 4X4 pixel segment. The pictures are mainly categorized into 3 types:
These pictures are characterized on the basis of how they coded. If we consider the I pictures then they use only information present in that frame only for coding where as the P- pictures uses information from nearest I or P pictures that are nearly present to it. The B- pictures uses information from both the past and future pictures. [beginersââ‚¬â„¢s guide]
Now for video encryption, there are two types of algorithms present total encryption and selective encryption algorithm.  We are using selective algorithm. In the selective encryption algorithm we do not encrypt all the data but we select only some part of data to encrypt. In the following section we will discuss the some of the parameters for encryption:
The elements considered are slice header, picture header, GOP header etc. among which the slice header, picture header, GOP header are not considered as strong candidate for the encryption as they have many drawbacks. But the Macroblock_type is considered as the good candidate for the encryption because it provides the best visual degradation effect.
We seen that video is consists of different types of frames. P and B frames uses the information from the I frames for coding. So I-frames are considered as the most important frames in the video sequence. When I-frames are changed then error propagates throughout the video sequence. The data contained in the I-frames is very huge so we cannot encrypt the entire frame. Hence we have to select such content that will give more visual degraded effect. If we consider the slice from the frames then the first Macroblock is considered as the most influential element for encryption. Because when first Macroblock is encrypted then error will propagate throughout that slice.
Hence we seen that I frames and Macroblock type are considered as the best elements for the selective encryption. Now in the following section we defined a architecture for video encryption in cloud environment also we will discuss how encryption is done.
System architecture for video encryption/Proposed Method:
Here we have proposed system architecture for video encryption in cloud environment.
Fig 1: general Architecture of Video transmission model
As described in the above proposed model architecture when any vendor or user(shown as sender in fig) uses cloud database to store its video data then the encryption service is provided to that vendor by the cloud service provider. Using that encryption service sender encrypt the data and then store in the cloud database. So it provides the secure transmission of data. When any user requested that video again using the decryption service video is to de decrypted and sent to the requester.
The encryption service is elaborated in the following way:
Fig 2: Encryption service
When any video is given as input to the encryption service then first audio file is extracted from it. Then the frames are extracted from the video. From this point our algorithm starts. It works as follows:
Encrypt every 100th or any other number of your choice Macroblock_type. Then according to the type of frame select the slice amount. If it is I or P frame then only compute the slice length. According to the frame type we will select amount of slice header to encrypt. If it is I frame then select largest five slices and if it is P frame then select largest 2 slices from that frame. If the Macroblock_type of selected frame is encrypted already then leave that frame and select the next largest frame. Then first 128 bits of each slice is encrypted using the AES or any other encryption algorithm. Then combine frames to make video and attach the extracted audio file with it and send it to the database to store.
Conclusion and Future work:
The organizations are shifting towards cloud more and more now days. This paper presents a video encryption model which helps organizations to keep their video data securely in the cloud database. Also the method used in this paper for video encryption is very effective. It encrypts very small amount of data from the video. Important factors such as macroblock_type, I and P frames are only encrypted. In this we achieved or reduced the ratio of encryption. Also AES algorithm is used to achieve more security. This method id mainly used for MPEG-2 compression format. We need to extend this for other formats also. Also when user asks for video then decrypted video is sent directly so there may chances of attack on data during transmission. So it need to decrypt at the user side but without adding overhead of decryption on the user. That is our future consideration.