모바일 環境에 適合한 效率的인 動的 그룹키 合意 프로토콜 設計

조석향 成均館大學校 2006 국내박사

The basic requirement for secure group communications over insecure public channels is that all group members must agree on a common secret key. This shared secret key, called the session key, can later be used to facilitate standard security services, such as authentication, confidentiality, and data integrity. Group key agreement protocols are designed to meet this requirement, with the fundamental security goal being to establish the session key in such a way that no one except the group members can know the value of the session key. In key agreement protocols, more than one party contribute information to generate the common session key. This paper presents contributory key agreement protocols in which the session key is derived as a function of contributions provided by all parties. Therefore in these contributory key agreement protocols, a correctly behaving party is assured that as long as his contribution is chosen at random, even a coalition of all other parties will not be able to have any means of controlling the final value of the session key. The mobile computing architecture is asymmetric in the sense of computational capabilities of participants. That is, the protocol participants consist of the stationary server (also called application server or service provider) with sufficient computational power and a cluster of mobile devices (also called clients) with limited computational resources. An asymmetric mobile environment is common in a number of applications such as Internet stock quotes, audio and music delivery, and so on. It is desirable to minimize the amount of computation performed by each group member in a group involving low-power mobile devices such as smart cards or personal digital assistants(PDAs). Futhermore we are required to update the group key with low computational costs when the members need to be excluded from the group or multiple new members need to be brought into an existing group. In this paper, we propose a dynamic group key protocol which offers computational efficiency to the clients with low-power mobile devices. We compare the total communicative and computational costs of our protocol with others and prove its security against an active adversary in the random oracle model. Also this paper presents a through performance evaluation of group key management technique which is suitable for pay-TV system by actual setup as the number of clients(subscribers) increases up to 100. Recently, the bilinear pairings such as the Weil and the Tate pairings defined on algebraic curves over a finite field have found applications in the design of cryptographic protocols. In this paper, we propose a pairing-based key exchange protocol that is a bilinear variant of the former protocol[32]. The proposed protocol achieves low communication complexity and provides some computational savings by the batch verification of signatures. We show that the security of our scheme is guaranteed against an active adversary in the random oracle model under the bilinear Diffie-Hellman(BDH) assumption.

Automatic Reverse Engineering for Clarification of Unknown Protocol Structure

Young-Hoon Goo 고려대학교 대학원 2020 국내박사

As the amount of Internet traffic increases due to newly emerging applications and malicious behaviors with their frequent updates, the amount of traffic that must be analyzed is also rapidly increasing. Many protocols that occur under these situations are unknown and non-documented. For efficient network management and security, a deep understanding of these unknown protocols is required. Although several protocol reverse engineering methods have been introduced in the literature, there is still no standardized method to completely extract a protocol specification. None of these methods infer abundant, detailed enough, or intuitive specification. Besides, all of these methods still face major limitations in automation and the genericity. All methods are designed only for a particular layer or type of protocols. Additionally, compared to research of various protocol reverse engineering methods, the research of how to verify the inferred results has received little attention. To combat such problems, in this thesis, we cover the three main aspects: the quality improvement of the inferred specification, the genericity of the method, an objective evaluation method. First, we propose a novel automatic protocol reverse engineering method to extract an abundant, intuitive, and clear protocol specification. The proposed method infers all the key elements of the protocol, which are syntax, semantics, and finite state machine (FSM), and extracts clear syntax by defining fine-grained field types and three types of format: field format, message format, and flow format. Second, we proposed a method that is designed to be generally applicable, regardless of the specific network environment and protocol type from the viewpoint of the TCP/IP reference model. Lastly, we introduce a method to verify the inferred results objectively and multifaceted evaluation metrics. We demonstrate the feasibility of the proposed method by applying it to several protocols in various layers of the TCP/IP reference model.

DBA protocol research for improving GPON bandwidth utilization

이봉춘 Graduate school, Korea University 2010 국내석사

GiGabit-Capable Passive optical network (GPON) is considered as an efficient point to multipoint solution to satisfy the increasing demand for bandwidth between service provider central offices (COs) and subscriber. A GPON is consists of an optical line terminal (OLT), optical network units (ONUs) and optical splitter. OLT located at the provider CO is connected to a number of ONU located at customer site through optical splitter. Since the GPON system is easy to install and maintenance costs are low, it can be powerful solution for FTTH system. GPON uses Time Division Multiple Access (TDMA) to avoid collision in upstream data transmission and Dynamic Bandwidth Allocation (DBA) for efficient upstream bandwidth management. Dynamic bandwidth allocation in GPON system is a significant issue for providing efficient and fair utilization of the upstream bandwidth while satisfying the QoS requirements of different traffic classes. DBA requires amount of bandwidth capacity to transmit the information for DBA execution. DBA is executed with some period to reduce this bandwidth. So this makes DBA Protocol assign fixed bandwidth for the data which is susceptible to delay. This thesis analyze GPON protocols and network configuration, the information about dynamic bandwidth allocation in GPON, a new DBA Protocol based on the transmission container-1 (TCONT-1) was proposed to use fixed bandwidth more efficiently, elaborated the specific protocol for improved scheme, and builds the GPON model for this protocol and test its performance. The simulation results show this protocol can improve the upstream bandwidth utilization of the GPON system.

Protocol Independent Multicast-Dense Mode를 통한 향상된 Internet Protocol 멀티캐스트 라우팅

최종원 京畿大學校 大學院 1997 국내석사

전통적인 네트워크는 두 호스트사이의 통신을 포함하는 Point-To-Point 방식이 대부분이었으나, 최근 통신망이 고속화되고 대역폭이 증가됨에 따라 LAN TV, CSCW(Computer Supported Cooperative Work), 화상회의(Video Conferencing), 분산 데이터베이스 시스템(Distributed Database System)등과 같이 네트워크에 다양한 요구가 이루어지고 있다. 이와 같은 요구에 따라 데미터를 전송하는 방법 중 멀티캐스트에 대한 연구 및 구현이 활발하게 이루어지고 있다. 본 논문에서는 멀티캐스트의 필요성, MAC Layer 멀티캐스트 및 IP 멀티캐스트에 대하여 살펴보고, IP 멀티캐스트의 여러 가지 방법 중 Protocol Independent Multicast(PIM)의 Dense Mede와 Sparse Mode의 동작을 연구한다. 이를 바탕으로 IP 데이터그램이 포함하고 있는 8비트와 Time To Live 필드를 이용한 알고리즘 및 각 라우터의 지역시간을 이용한 향상된 새로운 Dense Mode PIM을 제안한다. 또한 본 논문에서 제안한 방법 Dense Mode PIM과 기존 Protocol Independent Multicast와의 차이점을 서술하였으며, 기존의 방법으로는 적절히 적용할 수 없는 네트워크 환경에 제안한 방법을 적용시켜 송신지에서 수신지까지 효율적인 경로를 찾아내었다. Traditional network computing applications involve communication between two computers. However, some important applications such as LAN TV, Video conferencing, CSCW(Computer Supported Cooperative Work) and Distributed Database System require simultaneous communication between groups of computers. The simplest technique for multipoint routing calls for the data sender to build a set of point-to-point communications paths. However, this approach inherently requires the bandwidth-consuming duplication of data streams across any shared media in the path. Because bandwidth can be a limiting factor in the successful deployment of multimedia applications, this solutiion has been eclipsed by emerging techniques such as IP multicast. This paper describes the need of multicasting, multicast addressing, MAC layer multicast and IP layer muticast and explores Protocol Independent Multicast (PIM) Dense Mode and Sparse Mode operation that existing Internet routing protocol. I suggest that advanced internet protocol multicast routing algorithm use Time To Live(TTL) value of IP datagram header and use local time on router. And this paper describes feature of new algorithm. The existing IP multicast provided Protocol Independent Multicast(PIM) with two ways, Dense Mode PIM and Sparse Mode PIM, according to network environment. But the new algorithm of this paper operate all network environment with one routing algorithm on account of use one router. Also, this algorithm can find efficient path from sender to receiver.

재요청을 고려한 버스중재프로토콜의 성능 분석

김남수 漢陽大學校 大學院 1995 국내석사

무선 센서 네트워크 환경에서 가상 트리 주소를 이용한 효율적인 라우팅 프로토콜

이용훈 인하대학교 대학원 2014 국내석사

무선 센서 네트워크에서 사용되는 센서 노드는 배터리, 메모리 등의 제한된 자원을 가지고 있기 때문에 이를 효율적으로 사용하는 라우팅 프로토콜이 요구된다. 기존의 라우팅 프로토콜은 플러딩(flooding) 방식의 프로토콜로 전송 경로를 설정하기 때문에 에너지의 소모가 많다. 이를 해결하고자 노드에 주소를 할당하여 데이터를 전송하는 분산 해시 테이블 기반 라우팅 프로토콜이 최근 활발히 연구되고 있다. 이 방식은 플러딩을 사용하지 않아 네트워크의 자원을 적게 소모하지만, 주소를 표현하기 위해 메모리 자원을 사용하며 최단 전송 경로를 찾지 못할 수 있다는 단점이 존재한다. 본 논문에서는 플러딩을 사용하지 않으면서도 싱크 노드와 센서 노드 간에 최단 전송 경로로 데이터를 전송할 수 있도록 보장하는 라우팅 프로토콜인 ERP-VTA (Efficient Routing Protocol using Virtual Tree Address) 를 제안한다. 제안하는 프로토콜은 먼저 싱크 노드로부터 트리 형식으로 프로토콜에서 사용되는 두 가지 종류의 주소인 가상 주소와 보조 주소를 각 노드에 할당한 뒤 데이터 전송 시 데이터가 전송될 목적지의 노드의 주소를 이웃 노드의 주소와 비교하여 적절한 노드로 데이터를 전송함으로써 목적지 노드까지 데이터를 전송한다. 시뮬레이션 및 성능 분석을 위해 라우팅 프로토콜 시뮬레이터를 사용하여 본 논문에서 제안한 ERP-VTA 와 기존의 대표적인 분산 해시 테이블 기반 라우팅 프로토콜인 VCP (Virtual Cord Protocol) 및 VRR (Virtual Ring Routing) 의 성능을 비교하였다. 다양한 네트워크 토폴로지에서 전송 경로를 설정하는 성능을 비교하기 위해 각 프로토콜이 구하는 전송 경로의 길이를 비교하였고, 메모리 자원 사용량을 비교하기 위해 각 프로토콜이 데이터 전송 주소 값을 표현하기 위해 필요로 하는 비트 수를 비교하였으며, 마지막으로 데이터를 전송하는 성능을 비교하기 위해 데이터를 전송하는데 걸리는 지연시간을 비교하였다. 시뮬레이션 결과 ERP-VTA 프로토콜의 성능이 VCP 프로토콜에 비해 전송 경로 부분, 주소 표현 비트 수, 지연시간에서 각각 1.87배, 3.22배, 1.67배 향상되었다. VRR 프로토콜에 비해서는 각각 6.44배, 1.53배, 5.15배 향상되었다. In the wireless sensor networks, a sensor node has limited resources such as battery or memory. Hence, the routing protocol used in the wireless sensor networks is required efficient use of resources. Existing routing protocols determine a transmission path by flooding method which consumes a lot of energy. In order to solve this problem, distributed hash table based routing protocol is studied recently. This protocol reduces consumption of resources in the networks by using alternative method which allocates address to each node, not a flooding method. However, this method can’t find a shortest transmission path and requires additional memory resource to represent address. In this thesis, we propose the novel routing algorithm called Efficient Routing Protocol using Virtual Tree Address (ERP-VTA). The ERP-VTA guarantees the shortest transmission path even while using address allocation method rather than flooding method. ERP-VTA forms hierarchical addressing tree structure from the sink node. In this tree, each node has two addresses, virtual address and sub address. When a node transmits a data, the node compares destination address of the data with address of a neighbor node and then sends the data to the neighbor node which has the closest address from the destination. Performance analysis of the proposed routing algorithm is achieved using routing protocol simulator. We compare ERP-VTA with VCP (Virtual Cord Protocol) and VRR (Virtual Ring Routing) which representative distributed hash table based routing protocols. In the realistic network topology, we compare transmission path length of each protocol for the performance of the path finding, and then required bits to representative of address for the performance of memory usage, and last latency for the real transmission performance. Simulation result shows that ERP-VTA protocol improves than VCP protocol by 1.87 times in the transmission path, and 3.22 times in the address expression bits, and 1.67 times in the latency. Compare result by VRR is in each part, 6.44 times, 1.53 times, and 5.15 times improved.

Bluetooth protocol중 L2CAP (Logical link control and adaptation protocol) 계층 구현

김종현 홍익대학교 정보대학원 2001 국내석사

본 논문에서는 블루투스 프로토콜중 코어에 해당하는 L2CAP (Logical Link Control and Adaptation Protocol) 계층의 구현을 연구 한다. 근자에 들어 관심을 물러 일으키는 근거리 무선통신 중 가장 표준화 작업이 잘 되어있고 향후 5년 이내에 전세계 시장에 수억 개의 단말에 탑재 되리라 예상되는 블루투스를 아직은 어색하나 빠른 시간 내에 연구되어져 밀려드는 외국 기술부분과 견줄 수 있는 위치에 이르러야 하기에 이 연구를 시작한다. 과거에 SNA나 TCP/IP 그리고 X400 & X500과 같은 각종 통신 프로토콜의 엔진부문을 국내에 접목키 위해 업계에선 많은 비용을 투입하여, 도입 이식하는 작업을 반복하였다. 그러나 블루투스 스택은 연구 개발함으로써 불필요한 도움이 될 것이다. L2CAP의 연구 구현은 아주 작은 일부분에 지나지 않으며 앞으로 HCI , SDP, RFCOMM을 모두 완성한 완전한 프로토콜 스택을 연구 구현 할 것이다. 더 나아가 PROFILES 부분과 BASEBAND 및 LG/LM부분도 구현되어야 한다. 또한 이 모든 과정은 BQB 인증을 취득, 해외 유수 제품과 성능 및 기능을 견줄 수 있다. In this paper we describe a development of L2CAP (Logical Link Control and Adaptation Protocol) for Bluetooth Protocol. The Bluetooth is very good standardization specification among a short-distance Radio Communication that arouse people's interest, recently. Maybe the product using bluetooth will be made some hundred millions after 5 years. But we don't know about Bluetooth in detail and import product and technique from Europe & USA. So we must start to study R implement as soon as possible about this. A few years ago, we took part in migration & implement localization for SNA, TCP/IP & X400 engine that go to a lot of expense. This is not good job for a field of computer & communication in domestic. This is a motive that study & implement the L2CAP for Bluetooth protocol. This study is only a small & starting. we will be study & delvelopement about the other then Bluetooth Part HCI, RFCOMM, SDP, a lot of profiles, Baseband & LC/LM continually. Finally, we'd like to certify from BQB this study and be better as compared with abroad.

급성 흉통환자의 응급관리 알고리즘과 프로토콜 개발 : 급성 관상동맥 증후군 중심으로

조석희 연세대학교 간호대학원 2006 국내석사

응급실 업무 상황에 적합하고 과학적 원칙을 근거로 만들어진 실무지침은 일차적으로 진단이 내려지지 않는 상태로 생명의 위협을 당하고 있을지 모르는 환자를 관리 하는 응급실 간호사의 의사결정을 돕고 업무를 지원할 수 있는 유용한 지침이 될 수 있다. 흉통은 복통에 이어 응급실에 내원하는 가장 빈도 높은 주증상이며, 급성 관상동맥 증후군은 흉통 중 가장 빈도 높은 질환이다. 급성 관상동맥증후군의 조기 진단과 치료는 환자의 예후와 합병증에 직접적으로 관련되기 때문에 신속한 대처를 위한 다양한 의학적 실무지침이 개발되고 있으나 간호사의 의사결정이나 업무를 지원해주기 위한 실무지침의 개발 연구는 국내에서 시행된 바 없다. 따라서 본 연구는 흉통으로 내원한 환자를 빠르고 정확하게 사정, 평가하고 진단하여 적절한 간호 중재를 제공하기 위한 응급관리 알고리즘과 프로토콜을 개발하는 방법론적 연구이다.연구진행 과정은 문헌고찰로 알고리즘 경로와 프로토콜 항목을 도출한 다음 I 대학병원에 흉통으로 내원한 환자 중 급성 관상동맥증후군 15명과 비 관상동맥증후군 10명의 임상 기록지 분석을 통해 도출된 알고리즘과 프로토콜을 실무에 맞게 수정하여 예비 알고리즘과 예비 프로토콜을 확정하였다. 확정된 예비 알고리즘과 예비 프로토콜은 응급의학과 전문의와 전공의, 응급실 경력이 있는 간호학교수와 응급실 임상 경력 8년 이상 응급실 간호사로 이루어진 13명의 전문가 집단으로부터 타당도를 검증 및 수정 보완하는 과정을 거쳐 최종 알고리즘과 프로토콜을 완성되었다.본 연구의 결과는 다음과 같다.1. 문헌고찰을 통해 알고리즘 경로와 프로토콜 항목이 도출되었다. 알고리즘은 1차 사정과 관리, 혈역학적 불안정 환자의 구분과 관리, 2차 사정과 관리, 허혈성 흉통의 가능성 및 감별진단, ST분절의 상승 심근경색의 구분 및 관리와 ST분절의 상승이 없는 급성 관상동맥 증후군의 관리 및 급성 관상동맥 증후군의 가능성이 적은 저 위험군 환자의 구분과 관리, 입원 및 퇴원의 결정의 6단계로 이루어져 있으며 프로토콜은 알고리즘 각 단계에 따른 주 객관적 자료 수집, 검사 종류와 분석, 치료적 중재, 간호적 중재, 추후 관리 흐름에 대한 세부 항목 124개가 도출되었다.2. 임상 기록지 분석으로 문헌고찰에 의해 도출된 알고리즘의 세부 경로가 실무에 맞게 수정 되었고 예비 프로토콜 항목이 93개로 확정 되었다.3. 전문가 타당도는 알고리즘 총 평균 3.252, 프로토콜 총 평균 3.406으로 비교적 적절한 것으로 평가 되었다. 그러나 알고리즘의 4단계 평균이 2.846으로 낮은 점수를 보여 4단계를 중심으로 알고리즘의 전반적 흐름과 프로토콜의 세부 항목의 수정과 보완이 전문가의 의견에 따라 이루어졌다.4. 최종 수정된 알고리즘은 1차 사정 및 혈역학적 불안정 환자의 구분과 관리, 2차 사정 및 급성 관상동맥 증후군 잠정적 진단 관리, ST분절 상승 심근 경색과 ST분절의 상승이 없는 급성 관상동맥 증후군 관리, 입원 및 퇴원의 결정의 4단계로 최종 개발되었으며 프로토콜 항목도 102개 항목으로 최종 개발되었다.본 연구를 통해 개발된 급성 흉통환자 응급관리 알고리즘과 프로토콜은 간호사의 의사결정을 지원하고 간호 업무의 효율성을 높여 응급간호의 질 향상에 기여 할 것이며 임상에서의 지속적 수정과 보완은 보다 유용한 실무 지침으로써 발전을 또한 기대 할 수 있다. The clinical practice guideline prepared on the basis of the scientific principle suitable for the situation of the emergency department, can be a useful guideline that may help nurses in the emergency department who take care of patients whose lives are threatened even without receiving the primary diagnosis, make decisions and support their works. Chest pain is the second most frequent chief complaint to abdominal pain for patients who visit the emergency department, and acute coronary syndrome is the most frequent disease in chest pain. Since early diagnosis and treatment of acute coronary syndrome are very directly coupled with the convalescence, prognosis and complication of patients, various clinical practice guidelines for promptly coping with those symptoms have been developed. Nonetheless, there have been no researches about the development of practical affair guideline for supporting nurses’ decision-making or undertaking their duties in Korea. Therefore, this study is a methodological study that develops the algorithm and protocol of emergency treatment for offering appropriate nursing intervention by rapidly and accurately assessing, evaluating, and diagnosing patients who visit a hospital from chest pain.In order to proceed to the study, algorithm paths and protocol items were deduced from literature study, and then, the algorithm and protocol which were deduced from the analysis on the clinical record paper about 15 patients with acute coronary and 10 patients with non- acute coronary syndrome of those visited an I-university hospital from chest pain, were modified to practical affairs, and preliminary algorithm and preliminary protocol were completed The validity of the developed preliminary algorithm and preliminary protocol was verified and modified by 13 professionals that comprised doctors and residents of the department of emergency medicine, professors of nursing who experienced the emergency department, and nurses in the emergency department with more than 8 year’s clinical experience in the room. In that way, the final algorithm and protocol were developed. through the procedures of verifying and modifying, complement its validity from the professional group.The results of this study are as follows.1. Preliminary algorithm and protocol which was deduced by literature study algorithm comprised six steps including the first assessment and care, the classification and care of patients who are hemodynamically unstable, the second assessment and care, diagnosis of the possibility and discernment of ischemic chest pain, the classification and treatment of acute myocardial infarction with ST segment elevation(STEMI) and the care of acute coronary syndrome without ST segment elevation and the classification and care of patients in the low risk group of which possibility of acute coronary syndrome is low, and the decision of hospitalization and discharging from hospital. As for protocol, 124 items were selected as detailed items for the collection of subjective and objective data according to each step of algorithm, test type and analysis, therapeutic intervention, nursing intervention, and the follow-up care flow.2. Preliminary algorithm which was deduced by literature study was amended by the analysis on the clinical record paper and Preliminary protocol items were fixed as 93 detailed items3. The validity of professionals was evaluated to be moderately appropriate as it was proven 3.252 on the average for algorithm, and 3.406 on the average for protocol. However, the average of the four steps of algorithm posted a lower score of 2.846, thus, the modification of the overall flow of algorithm and details of protocol was made by focusing on the four steps according to professionals’ opinion.4. The final modified algorithm version was developed to have four steps including the first assessment and the classification and care of patients who are hemodynamically unstable, the second assessment and care of tentative diagnosis of acute coronary syndrome, the care of acute myocardial infarction with ST segment elevation and coronary syndrome without ST segment elevation, and the decision of hospitalization and discharging from hospital. Also, the protocol items were fixed to be the total 102 items.The algorithm and protocol for emergency intervention of acute chest pain patients, developed through this study might contribute to quality enhancement of emergency nursing by elevating nursing efficiency while supporting nurses’ decision-making, and continued modification and supplementation are expected to develop as a more useful guideline for practical affairs in clinical cases.

Road-aware and traffic-aware routing protocol in vehicular Ad hoc network

Pham, Thi Hong Graduate School, Korea University 2011 국내석사

The Vehicular Ad hoc Network (VANET) is one kind of mobile ad hoc networks that provides communications among nearby vehicles, and between vehicles and nearby fixed equipments. There are many applications, such as assisting driver with signage, warning dangerous roads, and telling the ways, etc., can be implemented in VANET. However, these applications require high packet delivery ratio, and low end-to-end delay. Since VANET has high dynamic topology and various vehicle densities, developing a routing protocol that can satisfy above requirements is still a big challenge. VANET routing simulation concept includes two aspects: one is the traffic simulation and the other is network simulation. The traffic simulation is performed to create traces of mobility mode. And the network simulator is used to evaluate network protocols and application in a variety of conditions. Beside the above two aspects, another important thing for position ? based routing protocol is location service. Location service enables a node to be aware of the location of the node that it wants to communicate with. This thesis studies on VANET routing protocol that includes VANET routing problem, mobility model for VANET routing evaluation, location services for positionbased routing protocol. Also, we present an efficient road and traffic-aware routing protocol (RTRP), in which the best path to transmit data packets is calculated based on distance and density factor in the 1st phase. This best path includes intersection sequent numbers that data packets would be transmitted along. By using the best path information, a greedy data forwarding algorithm is deployed in the 2nd phase on each road segment, based on a Reaching Intersection Time (RIT) and a Turning Direction Probability (TDP). We develop a mobility model that includes road intersection, traffic lights at intersections, various density areas, obstacles, etc., to validate the proposed routing protocol. Simulation shows that our developed mobility model is realistic and can adapt well to the proposed routing protocol. Besides, we also make a comparison among shortest path selection algorithm, road-aware path selection algorithm and density-aware path selection algorithm by using VanetMobisim and NS-2.33. The simulations results show that density-aware and road?aware algorithm outperforms traditional shortest path selection algorithm when the mobility and the number of connection are increased.

SPDY Protocol 적용에 따른 웹 페이지의 성능향상에 대한 연구 : 분산 컴퓨팅 환경에서

송준호 숭실대학교 정보과학대학원 2014 국내석사

1999년 HTTP/1.1 Protocol이 발표된 이후 약 15년간 웹 환경은 많은 변화를 겪었다. 사용자에게 다양한 기능을 제공하기 위해 웹 페이지 크기가 증가하였고 스마트 폰을 이용한 웹 페이지 요청은 이전과 비교할 수 없을 정도로 늘어났다. HTTP/1.1과 TCP(Transport Control Protocol)를 이용한 현재의 웹 통신 방식은 서버와 클라이언트의 연결이 단방향으로만 수행되어 요청/응답 속도의 지연이 발생하고 있다. 데이터의 전송을 담당하는 TCP 의 혼잡 회피(Congestion Avoidance) 성향으로 인해 웹 페이지의 리소스가 늦게 전송되는 문제가 발생 하고 있다. 또 전체 HTTP의 메시지 구조상의 한계로 인해 메시지 헤더의 내용은 압축되지 않고 클라이언트로 전송되고 메시지 바디의 내용만 선별적으로 압축되어 전송되고 있다. 이러한 문제들은 웹 서버에서 제공하는 웹 어플리케이션의 성능저하를 가져올 수 있다. 이러한 문제를 해결하기 위해 HTTP/2.0이 발표 될 예정이다. HTTP/2.0에 정식으로 포함될 예정인 Google의 SPDY(SPeeDY)는 IETF의 표준화 프로세스에 의해 연구되어 온 기술이다. SPDY는 TCP 기반의 전송계층을 그대로 사용하면서 HTTP/1.1의 데이터 전송 포맷 및 커넥션 관리부분을 보완한다. SPDY는 양방향으로 송신/수신이 가능한 HTTP 연결을 이용해 HTTP/1.1이 가지고 있는 단방향 전송에 따른 지연을 제거하고 응답속도 향상을 위해 응답 메시지의 우선순위를 부여한다. 또 메시지의 헤더 압축을 통해 요청/응답속도의 향상을 지향한다[1]. 즉 하나의 연결로 완전하게 다수의 요청/응답을 동시에 송/수신 하는 Protocol이라고 정의 할 수 있다. 실험적으로 SPDY Protocol을 적용한 웹 페이지의 경우 HTTP/1.1 기반으로 동작하는 SSL(Secure Socket Layer) 사용 웹 페이지와 비교해 볼 때 평균 약 33% 의 응답속도가 줄어드는 효과가 있다고 알려져 있다. 본 논문에서는 가상 머신을 이용한 분산 컴퓨팅 환경을 구축하고 웹 서버인 Apache를 분산 컴퓨팅 환경에 설치해 HTTP/1.1와 SPDY에서의 웹 페이지의 로딩속도를 비교한다. 이를 통해 SPDY에서 제공하는 HTTP Header의 압축과 Multiplex Stream이 웹 페이지 로딩 속도에 미치는 영향을 구체적으로 제시한다. 또 분산 컴퓨팅 환경에서 SPDY를 이용한 웹 어플리케이션의 최적화 된 동작 환경을 제안 할 수 있을 것이다. Since HTTP/1.1 Protocol released in 1999, Web environment has been changed almost 15 years. The size of the Web page increased for three times and the request to Web page via smartphones was incomparably explode. HTTP/1.1 and TCP (Transport Control Protocol), the current communication methods are using the one-way request / response connection which may occur delaying. Also responsible for congestion avoidance of the TCP could bring the delay of the web page load time. Due to these problems that provides web application can lead to performance degradation on a web server. The HTTP/2.0 will be announced to solve this problem. The HTTP/2.0 to be included Google's SPDY (announced SPeeDY) Protocol officially which developed by the IETF standardization process. SPDY is used the TCP as the transport layer in HTTP/1.1 and complement connection management and data transfer formats of HTTP/1.1 Protocol at the sametime. SPDY used HTTP connection which supports the bi-directional transmission / reception to remove transmission delay and to prioritize response message to improve response time of Web page. Also through the header of the message compression, response time could be shorten. That is, SPDY can be defined protocol that can manage the a large number of connections request / response at the same time completely. Experimentally, Web page base on SPDY Protocol loaded in browser faster than HTTP/1.1 based Web page with SSL(Secure Socket Layer) as approximately 33%. In this paper, using a distributed computing environment which composed of the Virtual Machine with Apache web server, compares loading time of web page in SPDY protocol to HTTP/1.1`s Using the features of offering HTTP Header Compression and Multiplexing-Stream in SPDY can explain impact of web page load time in detail. In addition, We can suggest optimized configuration in a distributed computing environment using SPDY.