Claim Missing Document
Check
Articles

Analisis Perbandingan Kinerja TCP Vegas Dan TCP New Reno Menggunakan Antrian Random Early Detection Dan Droptail Guntur Wahyu Pamungkas; Widhi Yahya; Heru Nurwarsito
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 10 (2018): Oktober 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (545.294 KB)

Abstract

Abstract Transmission Control Protocol (TCP) is a protocol on transport layer that have reliable properties in sending data. But reliable transmission can be distruped due congestion which are caused by increased traffic data. The problem can be solved by applying the correcrt TCP and queue management. In this research will be done simulation by comparing TCP Vegas and TCP New Reno by adding Random Early Detection and Droptail with scheme the addition of minthresh, max thresh and buffer capacity. After this research is done will get TCP and queue management that has the best performance in over coming congestion. The test results show, TCP Vegas is better than TCP NewReno with Random Early Detection, especially in packet delivery ratio, delay and packet drop. The results of packet delivery ratio test, TCP Vegas has the best performance when using Random Early Detection and Droptail with the average 99.9968% versus 99.6768% for TCP NewReno. The results throughput test, TCP NewReno has the best performance when using Droptail with the average 100749.4234kbps versus 94576.815kbps for TCP Vegas. The results delay test, TCP Vegas has the best performance when using Random Early Detection with the average of 4.31ms versus 5.98ms for TCP NewReno. And results packet drop test, TCP Vegas has the best performance when using Random Early Detection with the average 0.0002% versus 0.319% for TCP NewReno.
Implementasi Struktur Data Dictionary Untuk Sistem Monitoring Perangkat Internet Of Things Galeh Prehandayana; Widhi Yahya; Heru Nurwarsito
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 10 (2018): Oktober 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (629.404 KB)

Abstract

Development of internet has been increased in recent years. One of the well developed internet technology is called Internet of things (IoT). IoT is able to transmit data through the network without direct interaction of person-to person or person-to-computer. The increase in size and amount of IoT environments complemented with the complex usage oftenly cause disturbances on networks and devices. Well developed device management is required to observe and availability of networks to prevent disturbances. Microcomputers with Raspberry Pi is required to develop the system. The device monitoring system of IoT works on transport layer using UDP protocol, the system will take only the required data for effective computing. Data structure will be designed with Dictionary Data Structure system, the monitoring system of IoT then developed by modifying the Management Information Base (MIB). The development of agent and manager will be adjusted according to MIB on SNMP with OID categories to create a well structured information and data to create an accurate interface for device monitoring. The result of IoT device monitoring is processed from data package on one transmission with the size of 402 bytes, slighty different with previous research which is 326 bytes, the result is caused by the usage of dictionary data structure which key and value can choose data based on OID category. The IoT device monitoring system is more effective than 3916 bytes SNMP protocol.
Implementasi Metode Failover pada Broker Protokol MQTT Dengan ActiveMQ Mohammad Hafidzar Rakhman; Widhi Yahya; Kasyful Amron
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 10 (2018): Oktober 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (875.284 KB)

Abstract

Internet of Things (IoT) can be interpreted as a concept to connect "Things" or real objects in the world with the internet and sharing information. To share information between "Things" over the internet, Iot needs an ideal and efficient communication protocol in terms of resources when sending messages. One of the protocols considered to fit with IoT communication model is the Message Queuing Telemetry Transport (MQTT) protocol. In the MQTT protocol communication model, the broker has an important role in the success of the communication process that occurs. When the broker stops, the publisher and subscriber can't make the communication process and have to wait for the broker to be fixed. The solution of broker problem can be minimized by implementing failover between master and slave broker. Therefore, this study implements failover to improve the availability of services to brokers. Based on the test results, the system obtained an average downtime value of 24.3266 seconds from 5 tests. While on performance testing, the system gets an average latency value of 1.1763 seconds on the publisher side and 0.1157 seconds on the MQTT subscriber side. In packet loss testing, missing messages amount to 20 messages from 50 messages when implementing failoverl.
Analisis Perbandingan Performansi Kontroler Floodlight, Maestro, RYU, POX dan ONOS dalam Arsitektur Software Defined Network (SDN) Moh Wahyudi Putra; Eko Sakti Pramukantoro; Widhy Yahya
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 10 (2018): Oktober 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (324.446 KB)

Abstract

The main network infrastructure that is almost entirely managed by vendor leads to the increasing complexity of network built, so it is needed an approach to manage and implement the network to support the diverse needs, that is the needs of using Software defined network (SDN). Software defined network (SDN) has a basic concept of separating the control plane and data plane, so the network can be programmed individually from the data center. Some of the existing controllers are Maestro, Floodlight, RYU, POX and ONOS. To know which controller has the best performance, then doing a performance comparison analysis among the controllers. Tests conducted in this research is using mininet as an emulator to design test topology and Cbench to evaluate parameters of throughput and latency. Based on the test, we found that Maestro controller has better capability to handle data flow with larger amount of data compared to Floodlight controller, RYU controller, POX controller, or ONOS controller. Thus, the Maestro controller has a better chance to guarantee the processing of large amounts of data. With the difference of throughtput value in the range of 500 to 4000 flow/s. With Maestro controller, we also got better and superior latency values with the difference range of around 1000 ms up to 5000 ms.
Implementasi Link Fast-Failover Pada Multipath Routing Jaringan Software-Defined Network Muhammad Aji Wibowo; Widhi Yahya; Dany Primanita Kartikasari
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 10 (2018): Oktober 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (827.96 KB)

Abstract

Software-defined network is a concept of computer network approach where the control plane is separated by data plane. The communication between the data plane controller uses the OpenFlow protocol. In an SDN architecture it often happens to have interference such as failure link. in the latest openflow protocol there is a group type that overcomes the interruption of the link is using a fast-failover mechanism. Tests conducted to measure the performance of the system include testing functionality, response time and packet loss. The results of the test obtained from this study are able to search the path at the time before the link down and after the link down. At the time of link 1 in down in topology has a response time with an average of 0.38-0.681 ms and in link 2 has an average of 0.272 - 0.549 ms. When the link is down, the lost packet until it finds the new path can be known the value of packet loss. packet loss obtained from the simulation has an average of 0.423% -0.636%.
Implementasi Load Balancing untuk Kontroler Software Defined Network Adi Iman Utama; Widhi Yahya; Dany Primanita Kartikasari
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 11 (2018): November 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (496.464 KB)

Abstract

The development of increasingly complex needs of traditional networks will affect the high cost of investment and hardware maintenance so that required solutions to these developments by using Software Defined Network (SDN). SDN is a concept of separation between control plane and data plane on network devices such as routers and switches, whereas in traditional networks the control plane and data plane are combined into one. Using a single controller on the SDN concept obtained a latency value with an average of 1000 to 1900 responses / second and throughput with an average of 50 to 300 flow / second, the value decreased performance when the number of switches connected in the network environment increased. In this research apply load balancing to do the load sharing on some controller that connected so that performance of each controller can run with maximal. The controller used in this research is POX while to implement load balancing controller will be used HAproxy on server computer. The test results show that as the increasing number of controllers with load balancing will provide better latency and throughput results from the use of 3 controllers obtained the latency value with an average of 2000 to 3300 responses / second and throughput with an average of 70 to 4300 flow / second, but as the number of controllers increases load of server load balancing will increase seen from CPU and Memory Usage.
BitTorrent dengan Reactive Routing pada Software-Defined Network Landika Hari Suganda; Widhi Yahya; Mahendra Data
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 11 (2018): November 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (641.643 KB)

Abstract

BitTorrent is a file sharing application that uses a peer-to-peer mechanism to distribute pieces of files (chunk). The process of determining the seeder is still done randomly, so to get the optimal seeder needs to be optimistic unchoking. In this study BitTorrent is implemented with software a defined-network concept that separates the control plane and data plane to facilitate application development. The reactive routing controller creates a flow table that use to connect between the peers and determines the peer list as the nearest seeder based on the number of hop using the Dijkstra algorithm. The simulation results show that the system can select some of the nearest seeders by reactive routing based on the number of hops traversed. we use BitTornado for application BitTorrent client and use Pybtracker for application of tracker. Here we observe download rate and throughput generated on BitTorrent application with random seeder and BitTorrent application that has been modified with controller. The test results obtained an average value of download rate 4785.3 Kbps with 25 seconds distribution time on 3 random seeders, 6378.3 Kbps with a distribution time of 17 seconds on 5 random seeders, 6721.9 Kbps with 18 seconds distribution time on 7 random seeders and 5990.8 Kbps with time distribution of 18 seconds on 3 seeders, 6732.1 Kbps with 15 seconds distribution time on 5 seeders, 7656.53 Kbps with 15 seconds distribution time on 7 seeders.
Implementasi Multipath Routing Berbasis Algoritme DFS yang Dimodifikasi Uis Yudha Tri Wirawan; Widhi Yahya; Achmad Basuki
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 11 (2018): November 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1033.172 KB)

Abstract

Multipath routing is the routing technique that use some of the available paths. Multipath routing can reduce congestion on the network with load-balancing so that network traffic can be distributed to several available paths. It can be implemented with the concept of Software-Defined-Networking (SDN) which enable decoupling of control plane and data plane so as to facilitate the development of network applications. Here, we perform multipath routing implementation on the OpenFlow SDN and based on modified DFS and modified Djikstra as reference. The simulation results show that the system can perform multiple path search on a topology with the average response time of DFS algorithm is 612,7 ms. The average response time of modified-DFS algorithm is 632.6 ms and Dijkstra 325.6 ms. Number of iterations of the DFS algorithm requires 12694 iterations, modified-DFS requires 12860 iterations, and Dijkstra requires 800 iterations. The average execution time of DFS algorithm 0,0185 ms, modified DFS is 0,0258 ms, and modified Dijkstra 0,0005 ms. The result of throughput test shows that 50% of clients get throughput of 20 Mbps or less on modified-DFS algorithm. While at Dijkstra, as many as 50% of clients get throughput of 23 Mbps or less. The multipath test results show that load-balancing mechanism with group table still has unbalance distributed flow.
Implementasi Algoritme Blake2s pada JSON Web Token (JWT) sebagai Algoritme Hashing untuk Mekanisme Autentikasi Layanan REST-API Bagus Satria Wiguna; Ari Kusyanti; Widhi Yahya
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 12 (2018): Desember 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1106.597 KB)

Abstract

REST is a web-based client-server communication architecture. REST server provides data in the form of URL to be accessed by client in JSON form. The disadvantage of current REST architecture is the lack of authentication mechanism which allows anyone to access, change, or delete data. To solve the authentication problem from REST architecture, is required an authentication system. JWT is a token that is proposed to authenticate and ensure message integrity. Using JWT on the REST architecture can provide authentication and security. In JWT Implementation there are various hash algorithms used, one of them is HMAC-SHA256 algorithm which is SHA256 algorithm using message authentication code (MAC). In 2011, the vulnerability of SHA256 is found which is preimage attack and pseudo collision attacks. BLAKE2S algorithm is an algorithm created in 2012 succeding BLAKE algorithm. BLAKE2S algorithm offers better security since it is built with HAIFA iteration and ChaCha stream cipher. This research implemented BLAKE2S algorithm on JWT for REST API service authentication mechanism as alternative HMAC-SHA256 algorithm. Test-vector is performed to ensure the test-vector given in RFC document. The Authentication process of BLAKE2S is 88.583 ms.
Perbandingan Kinerja Hbase dan MongoDB Sebagai Backend IoT Data Storage Niki Yuniar Wicaksono; Eko Sakti Pramukantoro; Widhi Yahya
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 12 (2018): Desember 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (985.797 KB)

Abstract

IoT (Internet Of Things) is currently evolving as evidenced with many studies that discuss the IoT. IoT consists of two main components of the Internet and interconnected Things, to collect data and information into data storage media that can be processed as needed. However, there are challenges in building IoT data storage media, like a large data volumes, diverse forms and data formats. In previous research has developed a Framework of IoT data storage media from the sensor node. The Framework uses NoSQL MongoDB as its data storage. In this study, proposed an IoT NoSQL Hbase data storage using the MongoDB Framework because it free scema and good writing speed in the performance testing of random writing tests. This study, tested Hbase and MongoDB by testing ability to storage data in data storage. The result of this test is Hbase and MongoDB is able to receive and insert text data and image data. While for testing the Framework of data storage media with Hbase using storage testing For testing the performance of MongoDB, better result test on runtime parameters of 33s, throughput of 1559 ops/s, 48% Usage CPU, 67% Memory Usage, and 4354kb Disk I/O. When testing images the results of differences in slightly different better of MongoDB on Runtime parameters 64s, throughput 90 ops/s, CPU Usage 46%, Memory Usage 84%, while when testing Hbase images superior in Disk I/O parameters of 109.462kb.
Co-Authors Achmad Basuki Adhitya Bhawiyuga Adi Iman Utama Aditya Bhawiyuga Afredy Carlo Sembiring Agung Nugroho Ahmad Naufal Romiz Ahmad Rizal Muttaqin Andhika Dwitama Putra Andika Kurniawan Andreas Frederius Aprilia Kartika Sriastunti Aprillia Arum Pratiwi Ardy Frayogi ari kusyanti Arief Budi Setiawan Arnandha Hadi Pranata Aswin Suharsono Bagas Prawira Adji Wisesa Bagus Prasetyo Bagus Satria Wiguna Bhaskara Vito Tarigan Dany Primanita Kartikasari Dany Rahmana Dea Asmara Gita Delarta Tok Adin Dimas Dwi Radhityo Dimas Setiawan Afis Doni Hadiyansyah Enggar Saka Dirgantara Faizal Ramadhan Fariz Andri Bakhtiar Florensius Tri Putra Simamora Frondy Fernanda Ferdianto Gabreil Arganata Galeh Prehandayana Guntur Wahyu Pamungkas Hafidz Rizal Hafizhul Karim Hasbi Razzak Heru Nurwarsito Hilkya Dwima Palembangan Ibrahim Attamimi Imam Santoso Ismiarta Aknuranda Kasyful Amron Lalu Fani Islahul Ardy Landika Hari Suganda Lucky Ronny Chandra Negara M. Ali Fauzi Mahendra Data Maskur Purwiadi Moch. Wahyu Imam Santosa Mochammad Hannats Hanafi Mochammad Hannats Hanafi Ichsan Moh Wahyudi Putra Mohamad Rexa Mei Bella Mohammad Hafidzar Rakhman Muhammad Aji Wibowo Muhammad Luthfi Muhammad Rifqi Muzaki Muhammad Sholeh Muharrom Abdillah Niki Yuniar Wicaksono Nur Fauzi Panser Brigade Muhammad Poltak G. Hutajulu Pramukantoro, Eko Sakti Primantara Hari Trisnawan Rahadiyan Yuniar Rahmanda Rakhmadhany Primananda Ratih Kartika Dewi Reza Andria Siregar Rian Bayu Pambudi Risailin Dwi Jaka Fauzi Riski Julianto Romy Dwi Andika Manullang Sabriansyah Rizqika Akbar Sandy Ramadhan Wijaya Steven Urbani Tista Pamungkas Ragil Alit Uis Yudha Tri Wirawan Ulfa Khoirul Azizah Wijaya Kurniawan Yasin Rizqi Afandi Zavero Brillianata Abilovani Zefanya Hermanta