CAJ | 학술논문

为了应对爆炸式增长的车辆流量数据和用户日益增长的应用程序需求给蜂窝网络运营商(如3G/LTE)造成的巨大的流量负载和网络拥塞等问题，当前已有很多工作提出将数据通过高容量的用户设备间通信进行分流。通过蜂窝网络分发数据对象到一部分订阅用户(称为种子用户)，由种子用户通过本地机会通信的方式传送给其他订阅用户。为了最大化分流的效率，大多数的工作根据分流的对象和意义，提出了多种分流的方案，但是这些工作大都关注于种子用户的选择问题上。为了研究数据的缓存分配对分流的影响，着眼于城区VANET环境下的数据分流，根据城区多种数据异构的特性，提出利用公交车作为种子用户，通过制定数据缓存分配方案来间接完成数据分流的方案。缓存方案越好，用户命中自己想要的数据的概率就越大，分流的数据量就越多。据此，相应制定出几种贪心的缓存方案，一种启发式方案。通过大量的实验验证了启发式分配方案在分流数据量较多的情况下的优势，同时指出影响数据分流效率的几个重要参数。
위료응대폭작식증장적차량류량수거화용호일익증장적응용정서수구급봉와망락운영상(여3G/LTE)조성적거대적류량부재화망락옹새등문제，당전이유흔다공작제출장수거통과고용량적용호설비간통신진행분류。통과봉와망락분발수거대상도일부분정열용호(칭위충자용호)，유충자용호통과본지궤회통신적방식전송급기타정열용호。위료최대화분류적효솔，대다수적공작근거분류적대상화의의，제출료다충분류적방안，단시저사공작대도관주우충자용호적선택문제상。위료연구수거적완존분배대분류적영향，착안우성구VANET배경하적수거분류，근거성구다충수거이구적특성，제출이용공교차작위충자용호，통과제정수거완존분배방안래간접완성수거분류적방안。완존방안월호，용호명중자기상요적수거적개솔취월대，분류적수거량취월다。거차，상응제정출궤충탐심적완존방안，일충계발식방안。통과대량적실험험증료계발식분배방안재분류수거량교다적정황하적우세，동시지출영향수거분류효솔적궤개중요삼수。
The explosive growth of the vehicular traffic and the ever-increasing application demands of user cause huge traffic load, network congestion, and the pressure of the cost of maintenance for cellular network operators, such as 3G/LTE. To solve this problem, there have been many efforts to offload the traffic from cellular links to direct high-capacity device-to-device communication among users. The basic idea is to distribute the data object to only part of subscribers ( called seed sources) via the cellular network, and then allow seed sources to propagate the object to other sub-scribers through opportunistic local communications. In order to maximize the efficiency of offload-ing, most of works have studied the various forms of offloading scheme from the perspective of the objects and significance of offloading, but these works are mostly focused on the selection problem of seed sources. In this paper, a scheme is proposed, which uses the metropolitan move relatively sta-ble buses as seeds to develop buffer allocation scheme to indirect offload cellular traffic based on the characteristics of urban multiple heterogeneous data. The better buffer scheme, the greater probabil-ity that users hit the data they want, then the more data to offload. Therefore, several greedy buffer schemes and a heuristic scheme are proposed. Finally, the advantages of heuristic scheme which is used in the case of a large amount of data is verified by plenty of experiments, and several important parameters that influence the efficiency of the data offloading are pointed out.