考量惡意攻擊與天然災害下確保服務持續性之有效資源配置策略

Effective resource allocation strategies to assure service continuity considering malicious attacks and natural disasters

李佳玲(Chia-Ling Lee)

指導教授 ： 林永松

國立臺灣大學/管理學院/資訊管理學研究所/碩士(2013年)

https://doi.org/10.6342/NTU.2013.01258

摘要

現在各行各業或是政府幾乎都會使用網路提供服務給客戶，並且公司內部也會使用網路來作為內網的連結，公司內部許多的系統也是要用到網路。駭客就很常針對這些元件、系統來做攻擊，要是攻擊成功，常常會造成很大的損失。最直接的就是無法提供服務，間接的也會造成顧客的不信賴，甚至機密資料遭受竊取所造成的後果更加嚴重。因此每年公司還是會花很多的資金在防止網路攻擊造成的損失，尤其最近很盛行的協同攻擊，容易造成更大的破壞力，公司需要納入考量。
最近天災頻傳，尤其地震帶來的災害更是大面積的破壞，像是日本311大地震、四川大地震，都造成很大的損失，另外地震伴隨的火災也常造成更嚴重的損害。另外火災也常因為擴散性造成整個區域的燒毀，整個廠房常常就付之一炬，並且其發生的頻率更是高過大型的天然災害幾百倍，累加起來的損失也是不容忽視。
我們的目標就是希望能夠找出有效的防禦方法，來幫助企業在面臨惡意攻擊和天然災害仍能維持提供一定水準的服務，維持服務不中斷。本研究採用了在實務界很常使用的方法:redundancy來讓服務不間斷。
本研究使用數學規劃合併Monte Carlo Simulation來解決這個複雜、充滿randomness的問題，用模擬的方式來模擬整個network以更貼近現實，找到有效的防禦方法。並且也會持續做enhancement process以找到最佳的資源配置。

關鍵字

協同攻擊 ； 天然災害 ； 網路存活度 ； 數學規劃法 ； 模擬 ； 備援 ； 最佳化 ； 資源配置

並列摘要

Companies or governments rely on Internet to provide all kinds of service to customers and use Internet to propagate them in order to attract more customers to create more profits. Not only external customers, within the company, they also build their own intranet to handle daily operations. Once companies’ network being broken, they cannot provide regular service to user and also cannot run the daily process which may cause serious problem. Therefore, according to some research, cyber-attacks still is the most significant risk that business worried about since cyber-attacks will cause serious damage to company.
In addition, in recent decades, damage caused by natural disaster becomes more and more serious and happened more frequently than before. The number of disaster events reported globally increased from 1,690 to 3,886 and the economic losses also increase dramatically. Hence, in our thesis we want to add natural disaster this environment variable to our scenario. Companies need to start to pay attention on it when they are building their system. We discuss earthquake, secondary disaster-fire and fire in our scenario.
In order to provide business continuity, we also adopt redundancy this defense strategy to increase survivability. It is an effective method to prevent service interruption. When nodes damage or temporary shutdown, they can activate redundancy immediately which can prevent service interrupted. There are also other defense strategies to help defender maximize their system survivability such as virtualization, deploying honeypot, and cloud security
Our purpose is to help defender find out effective defense strategy and resource allocation. Our problem is a bi-level problem and we use mathematical programming combined Monte Carlo Stimulation to help us solve this complex problem since there are various of attack and defense strategies and full of uncertainty. Furthermore, we will do both commander and defender enhancement process in order to find out better solution.

並列關鍵字

Collaborative Attack ； Network Survivability ； Natural Disaster ； Secondary Disaster ； Redundancy ； Resource Allocation ； Optimization ； Mathematical Programming ； Stimulation

參考文獻

[1]Symantec, “2011 State of Security Survey”, 2011. http://www.symantec.com/content/en/us/about/media/pdfs/symc_state_of_security_2011.pdf

[3] “PlayStation Hackers May Have Stolen Data on 75 Million Users, Sony Says”, Cliff Edwards and Pavel Alpeyev, Apr 27, 2011.

[8] UNESCAP, UNISDR “The Asia-Pacific Disaster Report 2010”, The UN Office for Disaster Risk Reduction (UNISDR) and the UN Economic and Social Commission for Asia and the Pacific (ESCAP), October 2010

[9] UNESCAP, UNISDR “The Asia-Pacific Disaster Report 2012”, The UN Office for Disaster Risk Reduction (UNISDR) and the UN Economic and Social Commission for Asia and the Pacific (ESCAP), October 2012

[10] S. Xu, “Collaborative Attack vs. Collaborative Defense,” Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, 2009, Volume 10, Part 2, 217-228, 2009.