Wireless Cache Network Content Push Method Based on Non-Orthogonal Multiple Access Technology

Abstract

The present invention discloses a method for wireless cache network content pushing based on non - orthogonal multiple access technology. By means of the way that the channel gain serial numbers increase from back to front, the channel gain is taken as the benchmark for one iteration, the allocated power of each content cache file for each iteration is calculated, and the current pushing scheme is executed according to the allocated power of the iteration times that meet the power constraint. If the power pushing requirement is still not met in the last iteration, it is determined that the current pushing scheme does not meet the pushing requirement; the last content cache file in the current pushing scheme is removed, and the allocated power is readjusted to complete the dynamic pushing of the content cache file in the power domain. The present invention can save spectrum resources while improving the cache efficiency. In addition, the present invention maximizes the sum of the popularities of the successfully decoded content files in the cache server through a joint heuristic algorithm, obtains an approximate optimal solution with a low computational complexity, and lays a good foundation for the content delivery phase.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communication, and particularly relates to a method for pushing content in a wireless caching network based on non-orthogonal multiple access technology. Background Art

[0002] With the development of science and technology, people's requirements for service quality are getting higher and higher. As an important supporting technology for future communication networks, wireless caching pre-pushes content cache files to cache servers close to users during off-peak hours, so that subsequent user requests can be locally processed by the cache server, effectively reducing the system interruption probability and communication delay.

[0003] As another key technology in future networks, non-orthogonal multiple access technology can effectively alleviate the problem of the increasingly scarce radio spectrum resources. The main principle of non-orthogonal multiple access technology is to enable users to share time-domain or frequency-domain resources and distinguish users in the power domain. At the sending end, signals sent to different users are superimposed and power-distributed. At the receiving end, serial interference cancellation technology is adopted to preferentially demodulate the signal with stronger power, remove its interference after completion, and then demodulate the signal with weaker power.

[0004] By applying non-orthogonal multiple access technology to the content pushing stage of the caching network, multiple content cache files can be cached in the same time slot, greatly improving the caching efficiency. At present, certain achievements have been made in the research on applying non-orthogonal multiple access technology to wireless caching networks. In 2018, Z. Ding et al. proposed a transmission strategy for applying non-orthogonal multiple access technology to a wireless caching network in "IEEE Transactions on Communications", effectively combining the two technologies and studying the system performance. However, the scenario considered in this scheme is that the content popularity distribution may be different for different time slots, while in the real scenario, the demands of users in different regions are very likely to be different in the same time slot. Therefore, the caching efficiency and customer service quality of this scheme are not high. Summary of the Invention

[0005] In order to solve the above problems existing in the prior art, the present invention provides a method for pushing content in a wireless caching network based on non-orthogonal multiple access technology. The technical problems to be solved by the present invention are realized through the following technical solutions:

[0006] In a first aspect, a method for pushing content in a wireless caching network based on non-orthogonal multiple access technology is provided by the present invention. The method is applied to a base station, which communicates with multiple caching servers. The base station includes multiple content caching files whose popularity follows a Zipf distribution. Each content caching file has a one-to-one relationship with a caching server. The method for pushing content in the wireless caching network based on non-orthogonal multiple access technology includes:

[0007] S1. Obtain the channel gains between each caching server and the base station in the current time slot and the current pushing scheme, and sort the channel gains in descending order;

[0008] Among them, the current pushing scheme is a pushing scheme for pushing all content caching files to the corresponding caching servers in the current time slot. In each time slot, each content caching file corresponds to a decoding threshold;

[0009] S2. Assume that if the current pushing scheme is executed, in the way that the channel gain serial numbers increase from back to front, use the channel gain as the benchmark for one iteration, calculate the allocated power of each content caching file for each iteration, and compare the sum of the allocated powers with the total power to determine whether the allocated power of the current iteration meets the power constraint;

[0010] S3. If there is an iteration that meets the allocation constraint, execute the current pushing scheme according to the allocated power of this iteration. If the last iteration still does not meet the power pushing requirement, determine that the current pushing scheme does not meet the pushing requirement;

[0011] S4. Remove the last content caching file in the current pushing scheme as the current pushing scheme, and return to S2;

[0012] S5. Repeat S1 - S4 in each time slot to complete the dynamic pushing of content caching files in the power domain.

[0013] Optionally, the channel gains between each caching server and the base station in S1 are arranged in descending order as |h 1 | 2 >...>|h k | 2 >...>|h K | 2 , where |h k | 2 represents the k-th channel gain value after descending order arrangement; the current pushing scheme in each time slot is represented by f@{f 1 ,...,f n ,...,f N}, where f n= {0, 1} indicates whether the nth content caching file will be sent, where 0 means not sent and 1 means sent.

[0014] Optionally, S2 includes:

[0015] S21, if the allocated power in the (i - 1)th iteration does not satisfy the power constraint, then for the ith iteration, using the channel gain of the (K - i)th in descending order as a reference, calculate the allocated power of the (N - i)th content caching file;

[0016] where i ranges from 2 to K;

[0017] S22, starting from the (N - i)th content caching file, calculate the allocated power of each content caching file from back to front to obtain the allocated power of each content caching file in the ith iteration;

[0018] S23, compare the sum of the allocated powers of each content caching file in the ith iteration with the total transmission power of the base station to determine whether the ith iteration satisfies the power constraint;

[0019] S24, if the ith iteration does not satisfy the power constraint, then increment the value of i and return to S21.

[0020] Optionally, when the initial iteration i = 1, then using the last channel gain in descending order as a reference, the allocated power of the nth content caching file calculated is:

[0021]

[0022] where, |h k | 2 is the channel gain between the kth caching server and the base station, α n is the allocated power assigned by the base station to the nth content caching file, γ n is the decoding threshold of the nth content caching file, σ 2 is the Gaussian white noise power, and P is the total transmission power of the base station.

[0023] In S21 for the i = 1 iteration, if the allocated power is not satisfied, then start adjusting the power α of the last content caching file in the current push scheme last ;

[0024] where,

[0025] In S22, according to the adjusted α of the last content caching file in S21 last , update α from back to front last-1 , α last-2 ,..., α 1, the update principle is that for the previous content cache files, ensure that all cache servers can decode them normally. The update formula is:

[0026]

[0027] where n = last - 1, last - 2,..., 1.

[0028] Optionally,

[0029] The sum of the allocation powers of the i - 1 - th iteration's content cache files in S21 is If then it is determined that the i - 1 - th generation satisfies the power constraint; if then the allocation power of the i - 1 - th iteration does not satisfy the power constraint;

[0030] In S22, α adjusted according to the last - i - th content cache file in S21 last-i , update α from back to front last-i , α last-i-1 ,..., α 1 , the update principle is that for the previous content cache files, ensure that all cache servers can decode them normally. The update formula is:

[0031]

[0032] where n = last - i, last - i - 1,..., 1.

[0033] In a second aspect, a base station provided by the present invention implements the wireless cache network content pushing method based on non - orthogonal multiple access technology described in the first aspect.

[0034] The present invention has at least one or more of the following advantages:

[0035] 1. By introducing non - orthogonal multiple access technology into the wireless cache network and transmitting multiple content cache files in a superimposed manner, the present invention enables multiple content cache files to be pushed to the cache servers in the same time slot during off - peak hours, improving the caching efficiency.

[0036] 2. By using the joint heuristic algorithm to maximize the sum of the popularities of the successfully decoded content cache files in the cache servers, and caching the content cache files with larger popularities, that is, the content cache files with higher user request probabilities, to each cache server during the content pushing stage, the present invention lays a good foundation for the subsequent content delivery process and can improve the quality of service for users.

[0037] 3. The joint heuristic algorithm proposed by the present invention can obtain an approximate optimal solution with a lower computational complexity compared to traditional power allocation schemes.

[0038] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A diagram of a wireless cache network model based on non-orthogonal multiple access technology used in the present invention;

[0040] Figure 2 A general process diagram of a wireless cache network content push method based on non-orthogonal multiple access technology provided by the present invention;

[0041] Figure 3 A performance graph showing the variation of the total number of files in the file library with the popularity of the successfully received files of the present invention;

[0042] Figure 4 This is a performance diagram of the present invention showing that the sum of popularity of successfully received files varies with the number of cache servers. DETAILED DESCRIPTION

[0043] The present invention is further described in detail below with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

[0044] The present invention provides a wireless cache network content push method based on non-orthogonal multiple access technology, which is applied to a base station. Figure 1 As shown, the base station communicates with multiple cache servers, and the base station includes multiple content cache files whose popularity obeys Zipf distribution, and the content cache files and the cache servers are in a one-to-one relationship.

[0045] According to the characteristics of Zipf distribution, in order to ensure the current push plan, the base station will gradually make some cache servers unable to successfully decode the last content cache file to be sent in the plan. If all cache servers cannot decode the content cache file, the current plan will be abandoned. According to the idea of ​​greedy algorithm, the last content cache file will be chosen not to be sent in the push plan.

[0046] The present invention aims to propose a wireless cache network content push method based on non-orthogonal multiple access technology. Through the non-orthogonal multiple access technology, the content file information in the content file library is superimposed and transmitted in the power domain, thereby improving the spectrum efficiency while improving the cache efficiency and the user's service quality as much as possible.

[0047] like Figure 2 As shown, the present invention provides a wireless cache network content push method based on non-orthogonal multiple access technology, including:

[0048] S1, obtaining the channel gains between each cache server and the base station and the current push scheme in the current time slot, and sorting the channel gains in descending order;

[0049] Among them, the current pushing scheme is a pushing scheme that pushes all content cache files to the corresponding cache servers in the current time slot. Each content cache file corresponds to a decoding threshold within each time slot;

[0050] The channel gains between each cache server and the base station in S1 are arranged in descending order as |h 1 | 2 >...>|h k | 2 >...>|h K | 2 , where |h k | 2 represents the k-th channel gain value after descending order arrangement; the current pushing scheme in each time slot is represented by f@{f 1 ,...,f n ,...,f N}, where f n ={0,1} indicates whether the n-th content cache file will be sent, 0 indicates not to send, and 1 indicates to send.

[0051] S2. Assume that if the current pushing scheme is executed, in the way that the channel gain serial numbers increase from back to front, take the channel gain as the benchmark for one iteration, calculate the allocated power of each content cache file for each iteration, and compare the sum of the allocated power with the total power to determine whether the allocated power of the current iteration meets the power constraint;

[0052] In a specific implementation manner, S2 includes:

[0053] S21. If the allocated power of the (i - 1)-th iteration does not meet the power constraint, then for the i-th iteration, take the (K - i)-th channel gain in descending order as the benchmark and calculate the allocated power of the (N - i)-th content cache file;

[0054] Among them, the value range of i is from 2 to K;

[0055] S22. Starting from the (N - i)-th content cache file, calculate the allocated power of each content cache file from back to front to obtain the allocated power of each content cache file for the i-th iteration;

[0056] S23. Compare the sum of the allocated power of each content cache file for the i-th iteration with the total power to determine whether the i-th iteration meets the power constraint;

[0057] S24. If the i-th iteration does not meet the power constraint, then increment the value of i and return to S21.

[0058] It should be noted that: under the current push scheme, the base station first considers that all cache servers can successfully decode all the content cache files to be sent in this push scheme, and performs power allocation according to this situation.

[0059] If the k-th cache server can successfully decode the n-th content cache file, the following formula should be satisfied:

[0060]

[0061] where P is the total transmission power of the base station, |h k | 2 is the channel gain between the k-th cache server and the base station, α n is the power allocated by the base station to the n-th content cache file, γ n is the decoding threshold of the n-th content cache file, and σ 2 is the Gaussian white noise power.

[0062] Convert formula (1) into the form:

[0063]

[0064] It can be seen that if you want to satisfy the power constraint as much as possible then only the critical situation needs to be considered:

[0065]

[0066] Calculate the allocated power of each content cache file according to the above formula (3).

[0067] When the initial iteration number i = 1, based on the last channel gain arranged in descending order, the allocated power of the n-th content cache file calculated is:

[0068]

[0069] where |h k | 2 is the channel gain between the k-th cache server and the base station, α n is the allocated power of the base station to the n-th content cache file, γ n is the decoding threshold of the n-th content cache file, σ 2 is the Gaussian white noise power, and P is the total transmission power of the base station.

[0070] In S21, for the first iteration (i = 1), if the allocated power is not satisfied, then start to adjust the power α of the last content cache file in the current push scheme last ;

[0071]

[0072] It can be seen that α last is inversely proportional to |h k | 2 . Therefore, first consider the case where the cache server with the worst channel gain cannot successfully decode. According to the result of sorting by channel gain, that is, let the cache server corresponding to |h K | 2 not be able to successfully decode. At this time, α last is at most reduced to satisfy the following formula:

[0073]

[0074] In S22, α adjusted according to the last content cache file in S21 last , update α from back to front last-1 , α last-2 ,..., α 1 . The update principle is that for the previous content cache files, ensure that all cache servers can decode normally. The update formula is:

[0075]

[0076] where n = last - 1, last - 2,..., 1.

[0077] The sum of the allocation powers of the content cache files in the (i - 1)-th iteration in S21 is If then it is determined that the (i - 1)-th generation satisfies the power constraint; if then the allocation power in the (i - 1)-th iteration does not satisfy the power constraint;

[0078] In S22, α adjusted according to the (last - i)-th content cache file in S21 last-i , update α from back to front last-i , α last-i-1 ,..., α 1 . The update principle is that for the previous content cache files, ensure that all cache servers can decode normally. The update formula is:

[0079]

[0080] where n = last - i, last - i - 1,..., 1.

[0081] S3. If there is an iteration that satisfies the allocation constraint, then execute the current push scheme according to the allocation power of that iteration. If the last iteration still does not meet the power push requirement, then it is determined that the current push scheme does not meet the push requirement;

[0082] It should be noted that: if the i-th iteration satisfies the allocation constraint, the current push scheme is executed according to the allocated power of this iteration, so that the base station performs power allocation according to the current push scheme and sends the content cache file to the corresponding cache server.

[0083] S4. Remove the last content cache file in the current push scheme as the current push scheme, and return to S2;

[0084] It should be noted that: if the last iteration still does not meet the power push requirement, it is determined that the current push scheme does not meet the push requirement, and the push scheme should be adjusted, let f last = 0, where last refers to the position of the last content cache file in the current push scheme, and go to S2.

[0085] S5. Repeat S1 - S4 in each time slot to complete the dynamic push of the content cache file in the power domain.

[0086] The present invention discloses a method for pushing content in a wireless caching network based on non - orthogonal multiple access technology. By the way that the channel gain numbers increase from back to front, the channel gain is used as the benchmark for one iteration, calculate the allocated power of each content cache file for each iteration, and execute the current push scheme according to the allocated power of the iteration that meets the power constraint. If the last iteration still does not meet the power push requirement, it is determined that the current push scheme does not meet the push requirement; remove the last content cache file in the current push scheme and readjust the allocated power to complete the dynamic push of the content cache file in the power domain. The present invention can save spectrum resources while improving the caching efficiency. In addition, the present invention maximizes the sum of the popularities of the successfully decoded content files in the cache server through a joint heuristic algorithm, obtains an approximate optimal solution with a low computational complexity, and lays a good foundation for the content delivery phase.

[0087] The present invention provides a base station to implement a method for pushing content in a wireless caching network based on non - orthogonal multiple access technology provided by the present invention.

[0088] The performance of the present invention will be further described below in combination with simulation experiments.

[0089] Figure 3 A performance graph of the total popularity of the successfully received files of the present invention changing with the total number of files in the file library is given. Figure 4The performance graph showing the change in the total popularity of successfully received files of the present invention with the number of cache servers is given. The global adjustment is the power allocation adjustment strategy in the content push scheme of the present invention, and the local adjustment refers to only locally adjusting the power of the last content cached file in step S4. It can be seen that when changing the total number of files in the file library or the number of cache servers, the results of global adjustment are better than those of local adjustment. For Figure 3 , the total popularity of successfully received files for both schemes decreases to a certain extent. This is due to the characteristics of the Zipf distribution. Since the sum of popularities is 1, the more the total number of files, the lower the popularity of each file. Given a certain total transmission power, the number of files that can be selected does not change significantly. Therefore, the approximate optimal value of the objective function generally shows a downward trend.

[0090] Although the present application has been described in conjunction with various embodiments, however, in the process of implementing the claimed present application, those skilled in the art can understand and achieve other variations of the disclosed embodiments by viewing the accompanying drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "one" does not exclude a plurality.

[0091] The above content is a further detailed description of the present invention in combination with specific preferred embodiments. It cannot be determined that the specific implementation of the present invention is limited only to these descriptions. For those of ordinary skill in the technical field to which the present invention pertains, without departing from the concept of the present invention, several simple deductions or substitutions can be made, which should all be regarded as belonging to the protection scope of the present invention.

Claims

1. A method for wireless cache network content pushing based on non - orthogonal multiple access technology, which is applied to a base station. Characterized in that, The base station communicates with multiple cache servers. The base station includes multiple content cache files whose popularity follows Zipf distribution. There is a one - to - one relationship between the content cache files and the cache servers. The method for wireless cache network content pushing based on non - orthogonal multiple access technology includes: S1. Obtain the channel gains between each cache server and the base station and the current pushing scheme in the current time slot, and sort the channel gains in descending order. Wherein, the current pushing scheme is a pushing scheme for pushing all content cache files to the corresponding cache servers in the current time slot. In each time slot, each content cache file corresponds to a decoding threshold. S2. Assume that if the current pushing scheme is executed, then, in the order of increasing channel gain serial numbers from the back to the front, use the channel gain as the benchmark for one iteration, calculate the allocated power of each content cache file for each iteration, and compare the sum of the allocated powers with the total power to determine whether the allocated power of the current iteration meets the power constraint. S3. If there is an iteration that meets the allocation constraint, execute the current pushing scheme according to the allocated power of this iteration. If the last iteration still does not meet the power pushing requirement, it is determined that the current pushing scheme does not meet the pushing requirement. S4. Remove the last content cache file in the current pushing scheme as the current pushing scheme, and return to S2. S5. Repeat S1 - S4 in each time slot to complete the dynamic pushing of content cache files in the power domain. The channel gains between each cache server and the base station in S1 are arranged in descending order as |h 1 | 2 >...>|h k | 2 >...>|h K | 2 , where |h k | 2 represents the k-th channel gain value after descending order arrangement; the current push scheme in each time slot is represented by , where f n ={0, 1} indicates whether the n-th content cache file will be sent, 0 means not sent, and 1 means sent; S2 includes: S21. If the allocated power of the (i - 1) - th iteration does not meet the power constraint, then for the i - th iteration, use the (K - i) - th channel gain in the descending order as the benchmark to calculate the allocated power of the (N - i) - th content cache file. Wherein, the value of i ranges from 2 to K. S22. Starting from the (N - i) - th content cache file, calculate the allocated power of each content cache file from the back to the front to obtain the allocated power of each content cache file for the i - th iteration. S23. Compare the sum of the allocated powers of each content cache file for the i - th iteration with the total transmission power of the base station to determine whether the i - th iteration meets the power constraint. S24. If the i - th iteration does not meet the power constraint, increment the value of i and return to S21. When the initial iteration number i = 1, the allocated power of the n - th content cache file calculated using the last channel gain in the descending order is: where |h k | 2 is the channel gain between the k-th cache server and the base station, α n is the allocation power assigned by the base station to the n-th content cache file, γ n is the decoding threshold of the n-th content cache file, σ 2 is the Gaussian white noise power, and P is the total transmission power of the base station; In S21, for the i = 1st iteration, if the allocated power is not satisfied, start adjusting the power α of the last content cache file in the current push scheme last ; α adjusted according to the last content cache file in S21 in S22 last , update α from back to front last-1 , α last-2 ,..., α 1 , the update principle is that for the previous content cache files, ensure that all cache servers can decode normally, and the update formula is: Where n = last - 1, last - 2,..., 1; The sum of the allocation powers of the content cache files in the (i - 1)-th iteration in S21 is If then it is determined that the (i - 1)-th generation satisfies the power constraint; if then the allocation power in the (i - 1)-th iteration does not satisfy the power constraint; α adjusted according to the last-i content cache file in S21 in S22 last-i , update α from back to front last-i , α last-i-1 ,..., α 1 , the update principle is that for the previous content cache files, ensure that all cache servers can be normally decoded, and the update formula is: Where n = last - i, last - i - 1,..., 1.

2. A base station Characterized in that, It implements the method for wireless cache network content pushing based on non - orthogonal multiple access technology described in claim 1.

Images