USB storage devices can instantly copy large amounts of private data from your computer without installing any drivers. However, it is easy to cause computer data leakage, which is caused by the small storage capacity of MTU value and slow transmission speed. The new security problem of USB is prominent, and the optimization technology has been put forward at home and abroad. A common optimization technique is to use software to set the MTU value to control the flow of data. But the control software is easy to be hit by the virus work, can not effectively control the data flow leakage. Therefore, the research on MTU value configuration optimization technology of USB storage device is presented.

0 the introduction

USB storage devices can instantly copy large amounts of private data from your computer without installing any driver. However, it is easy to cause computer data leakage, which is caused by the small storage capacity of MTU value and slow transmission speed. The new security problem of USB is prominent, and the optimization technology has been put forward at home and abroad. A common optimization technique is to use software to set the MTU value to control the flow of data. But the control software is easy to be hit by the virus work, can not effectively control the data flow leakage. Therefore, the research on MTU value configuration optimization technology of USB storage device is presented.

USB storage device MTU value configuration optimization technology is mainly responsible for strengthening the privacy and real-time data transmission. Traditional technologies are still immature in terms of data availability. The core problems are: first, when users use USB storage devices, they may cause privacy leakage, which is not allowed in cases of better security. Second, when the user is offline, the data stored is completely unavailable. Without targeted technology, data availability can be reduced. In order to prevent the disclosure of private data, it is necessary to set the MTU value to configure the total storage capacity, improve the availability of data, and achieve the goal of load balancing.

1 behavior data collection

inserts a data acquisition code in the access control module of USB storage device, and sets two collection points at the entrance and exit of system call. One point is the logical end of the host and the other point is the logical end of the device flow.

collection process; On the USB bus, collect behavior data by bit. Define the data collection type with the package as the basic unit. When a data collection is completed between the host and the device flow, different types of packages are needed to check whether the collection is successful. The types include IN thing package, OUT thing package, and SETUP thing package. When it is collected, two packets with a specific format are applied, a transaction is started and sent to the host.

Descriptor refers to the relationship between the device and the USB storage device. Enumeration procedure :

Step1: after the USB host detects the device is inserted, the device shall be reset. The address after reset is 0;

Step2: get the device descriptor. In this process, the host can only read one packet, and the descriptor of the packet is 15 sections. If the breakpoint of USB storage device is 0, the device descriptor is less than 15 sections. At this time, the device descriptor should be returned according to the actual size of the breakpoint of 0.

Step3: reset again, set the address, the default address is 0;

Step4: when the host obtains the descriptor again, a new descriptor can be obtained from the new address of the host using the device. If the USB storage device endpoint is 0, multiple input requests are required;

Step5: the host obtains the configuration descriptor of the USB storage device, with a length of 8 bytes;

Step6: obtain a similar configuration set according to the length of the multi-description configuration set;

Step7: get the MTU value;

Step8: end of enumeration.

2. Total storage capacity setting

stores the collected behavior data to the MTU value configuration, and USES erasure correction code technology to select the appropriate total storage capacity. Let each behavior data create n identical copies, and at the copy node, establish the relationship model with probability theory.

after calculating the probability of online node and candidate node being unavailable at the time of h, it is necessary to detect whether the copy of behavior data meets the following conditions when selecting the number of storage nodes :

A: each behavior data has at least two copies, in probability calculation,f must be greater than 2.

B: data availability maintenance operations need to be completed in a certain amount of time. If the specified time is exceeded, the maintenance fails.

after finding the copy node of each data, the erasure correction technique is used to divide the behavior data into m data fragments. Will x copy code stored in the MTU value configuration, the remaining ( n - x ) is deposited in the cloud server.

to ensure that each behavior data copy node is of the same order of magnitude, the MTU value is set to configure the total storage capacity according to the following criteria :

first: at the current moment, the MTU value configuration must be online.

second: at the current moment, the MTU value configuration needs to have free resources.

Third: in the future h period,MTU value configuration needs to maintain data availability.

at a given time of h, the candidate nodes that do not meet the condition are eliminated. Using the grouping storage technology, the behavior data is divided into different groups according to the size. Each file can only be saved in one group, and the behavior data ID consists of a relative path and a data server. However, due to the large amount of data, an extant technique similar to the traditional technique is proposed to reduce the number of address Pointers and the total amount of behavior data.

in the MTU value configuration, the file presents a hierarchical tree structure.

, indexed by Indole, stores each file in a fixed location on the partition. Indole records the properties of each file, mainly including the file size and creation time, as well as the physical storage location of the behavior data on the MTU value configuration. The USB storage device behavior data is chunked in a contiguous space of 1 kB.

3. Configure space optimization

first collects the behavior data of USB storage device, and then sets the MTU value to configure the total amount of storage. In order to optimize the storage data, the flat storage structure is adopted.

in the USB data area, there is a lot of behavior data. Each behavior data area contains an Imsex Slot. Imsex slots represent behavior data in the stored data area. When a USB storage device stores small data, an index entry needs to be created and saved to an Imsex Slot. In order to reduce the amount of data, each Imsex Slot contains only 2 fields, the data size is 3 bytes, the physical address is 7 bytes, and other data is saved to the header. Finally, set the data block size and the average data size, if the data area space is exhausted, need to open up a new storage area in the remaining space.

it can be seen from figure 3 that all fragments will be connected into a fragment chain, and the fragment address placed in the head will be saved in a fixed position in the Imsex Slot.

when adding a new data, to be appended to the end of the space. The data area pointer moves toward the end of the currently available data area, in which case the data server continues to add new behavior data to the original data block. If the storage space is exhausted, it is necessary to find the appropriate data area to store the new data through the fragmentation chain.

from the optimization process, you can see that in the case of running out of storage space, you need to find the new most matching fragment and delete the data in the fragment chain number.

based on the above steps, optimization technology, optimization model.

optimization step: first, from the USB storage device, query the size of the access data MTU value configuration. Second, access the data directly on the MTU value configuration.

on access, the USB storage device will first update the memory data. If the memory data does not match the MTU value configuration data, there is an exception on the server. After the server is restarted, a consistency policy is set if the data loaded by the MTU value configuration is not the latest version of the data. Policy contents are :

A: the USB storage device will store the behavior data and corresponding data into the MTU value configuration, update the memory data, and return the operation results to the upper layer and the application.

B: there is no need to change the metadata during the operation, otherwise it will cause data consistency problems.

C: all operations can only be done in memory.

from this, complete the USB storage device MTU value configuration optimization technology research.

4. Experimental analysis

4.1 Experimental environment

Application Dell PC, Intel Q8200 4 core processor, 500 GB SATA hard disk. USB storage device partition size is 128 GE. Use the release 5.5 operating system and install the e4fsprogs software package.

4.2 The experimental requirements

Data sizes range from 1 kB to 64 kB, creating 2 048 data block files with a size of 64 kB. In a USB storage device, seven partitions are used and the size is 16 kB. In the experiment, the partition size should be the same.

4.3 The experimental steps

Step1:Within the specified data partition, a data block is selected。

Step2:Reads a certain length of behavior data from the offset。

Step3:Repeat Step1 and Step2 for 1 500 times。

4.4 The experimental results

(1)Random read and mixed read characteristics

based on the above experimental platform, the random read performance and mixed read performance of traditional optimization technology and MTU value configuration optimization technology of USB storage device were tested, and the results were compared.

analysis figure 5 shows that the performance of the built technology is 35% higher than that of the traditional technology, which is closer to the baseline performance, so as to achieve the approximate optimal. From the experimental comparison results, it can be seen that with the increase of data size, the performance advantage of USB storage device MTU value configuration random read increases. This is because the USB storage device MTU value configuration can optimize data access performance and reduce data addressing overhead. As the data size increases, the overall data access overhead ratio decreases, and the advantages of MTU value configuration optimization technology for USB storage devices become more obvious.

in figure 6,98:2 means that 80% of the mixed reads and 1% of the writes. Similarly,96:4 means that 90% of the random reads are read and 3% of the writes.

The analysis of figure 6 shows that the MTU value configuration of the USB storage device can achieve approximately optimal performance, which is far superior to the mixed read performance of the traditional configuration. In the experiment, it was found that as the data size increased, the performance advantage of using USB storage device MTU value configuration for mixed reads became more and more obvious, which was consistent with the test results in figure 5.

in summary, compared with the performance of USB storage device MTU value configuration for random read and mixed read, the traditional configuration is faster, indicating that the optimization technology of this design is more practical.

(2)Data consistency strategy

USES strong consistency strategy and weak consistency strategy to analyze their influence on the performance of the two technologies.

As can be seen from figure 7, in this experiment, all data were randomly selected from the database. When using a weak consistency policy, the proposed optimization technique increases with the proportion of write requests in total requests. Explanation for this result: under the weak consistency policy, the write operation writes the data back to the system, and the system periodically writes the new data to the storage. When using the strong consistency strategy, the write performance of the optimization technique will improve with the increase of the write operation proportion. This is the experimental result of synchronizing the new data directly to disk when using the strong consistency strategy.

traditional optimization techniques in the use of strong consistency strategy, with the increase in the proportion of write operations will decline, this is because the probability of hitting data in the cache is too low, resulting in a disk operation, resulting in time spent more than the write operation. With a weak consistency policy, the proportion of write requests in total requests decreases with each pass, and the problem with this is that when a write puts the data into the cache, the disk cannot read it, which degrades the performance of the write operation.

From the experimental results, it can be seen that no matter which strategy is used, the performance of MTU value configuration optimization technology of USB storage device is better than that of traditional optimization technology.

5 conclusion

user privacy data availability has always been the focus of research at home and abroad, for this reason, in view of the problems of traditional optimization technology, USB storage device MTU value configuration optimization technology. In theory, users who want to delete historical private data need to migrate the historical data to the MTU value configuration. In order to improve the availability of data, MTU value was adjusted to configure the total capacity by using erasure correction code technology to achieve the expected data availability. In order to reduce the impact of space opening, a fragmentation management strategy is designed to help the original data block find a suitable data area to store the new data. Finally, based on the above design steps, an optimization model is established to enhance the security of data access. In order to verify the feasibility of MTU value configuration optimization technology for USB storage device, the experimental environment and requirements were set in the experiment. In this platform, the performance of the proposed optimization technique for mixed and random reads is tested. According to the experiment, with the increase of data volume, the more obvious the advantages of the proposed optimization technology are, the higher the feasibility is. However, it is found in the experiment that the MTU value configuration optimization technology of USB storage device is not very stable in the process of repeatedly reading behavioral data, and it needs to be further improved.