scispace - formally typeset
Search or ask a question

Showing papers by "Telkom Institute of Technology published in 2010"


Proceedings ArticleDOI
17 Nov 2010
TL;DR: This paper designs a multi-path protocol which will use the same weighing method on BMC and a few modification on weight calculation by exploiting the Bandwidth Estimation on Westwood+ congestion control, this protocol will be mentioned as Normalize Uniform High-Adaptability (NUHA) TCP.
Abstract: Multi-path is a technique to increase the reliability of the internet connection in IP network. The implementation of this technique can be described as a host with two TCP paths work together, when one of them fails the other will back up the connection. Since TCP does not support multi-homing, the right term for this technique is multi-path. The fairness problem arises when multi-path host uses the same bottleneck link with another single-path host. This problem can be solved by scaling down each sub-flow on the multi-path by modifying the Congestion Control on Westwood+. This paper will design a Multi-path protocol which will use the same weighing method on BMC and a few modification on weight calculation by exploiting the Bandwidth Estimation on Westwood+ congestion control, this multi-path protocol will be mentioned as Normalize Uniform High-Adaptability (NUHA) TCP. The simulation will be conducted on HSDPA air interface to represent the wireless condition using network simulator 2. The results show that the multi-path protocol NUHA TCP works well in high loss network without violating the fairness.

7 citations


Posted Content
TL;DR: A mathematical model of quantum transistor in which bandgap engineering corresponds to the tuning of Dirac potential in the complex four-vector form is proposed, and simultaneous transformation of qubit and energy can be accomplished by setting the {control, cyclic}- operators attached on the coupling between one-qubit quantum gate.
Abstract: We propose a mathematical model of \textit{quantum} transistor in which bandgap engineering corresponds to the tuning of Dirac potential in the complex four-vector form. The transistor consists of $n$-relativistic spin qubits moving in \textit{classical} external electromagnetic fields. It is shown that the tuning of the direction of the external electromagnetic fields generates perturbation on the potential temporally and spatially, determining the type of quantum logic gates. The theory underlying of this scheme is on the proposal of the intertwining operator for Darboux transfomations on one-dimensional Dirac equation amalgamating the \textit{vector-quantum gates duality} of Pauli matrices. Simultaneous transformation of qubit and energy can be accomplished by setting the $\{\textit{control, cyclic}\}$-operators attached on the coupling between one-qubit quantum gate: the chose of \textit{cyclic}-operator swaps the qubit and energy simultaneously, while \textit{control}-operator ensures the energy conservation.

6 citations


Proceedings ArticleDOI
01 Oct 2010
TL;DR: A comprehensive design simulation of hybrid network concerning the timeout of ACK and CTS, which determine the maximum length of optical fiber is proposed and the result of simulation is used to predict the delay bound of inter-devices WLAN.
Abstract: The convergence between WLAN and optical fiber has created a new challenge in network design. In this paper, we propose a comprehensive design simulation of hybrid network concerning the timeout of ACK and CTS, which determine the maximum length of optical fiber. Then, the result of simulation is used to predict the delay bound of inter-devices WLAN.

5 citations


Journal ArticleDOI
TL;DR: The Ramsey number R(L, 2Km), where L is a linear forest is obtained, which is an extension for R(kPn, Hm) proposed by Ali et al.
Abstract: Let H be a graph with the chromatic number h and the chromatic surplus s. A connected graph G of order n is called H-good if R(G, H) = (n - 1)(h - 1) + s. In this paper, we show that Pn is 2Km-good for n ≥ 3. Furthermore, we obtain the Ramsey number R(L, 2Km), where L is a linear forest. Moreover, we also give the Ramsey number R(L, Hm) which is an extension for R(kPn, Hm) proposed by Ali et al. [1], where Hm is a cocktail party graph on 2m vertices.

5 citations


Book ChapterDOI
19 Jul 2010
TL;DR: The Ramsey number R(L, 2Km), where L is a linear forest is obtained, which is an extension for R(kPn, Hm) proposed by Ali et al.
Abstract: Let H be a graph with the chromatic number h and the chromatic surplus s. A connected graph G of order n is called H-good if R(G,H) = (n - 1)(h - 1) + s. We show that Pn is 2Km-good for n ≥ 3. Furthermore, we obtain the Ramsey number R(L, 2Km), where L is a linear forest. In addition, we also give the Ramsey number R(L,Hm) which is an extension for R(kPn, Hm) proposed by Ali et al. [1], where Hm is a cocktail party graph on 2m vertices.

3 citations


Proceedings Article
01 Aug 2010
TL;DR: Computer simulations showed that this method has performance better than the original VAD G729B in all tested noise conditions, and this method is improved in voice activity detection using K-Nearest Neighbor method for voice or unvoiced classification.
Abstract: ITU-T G.729B describes about Discontinuous Transmission (DTX) in the system. Discontinuous Transmission (DTX) is known as a transmission method that transmits a few bits of voice under influenced of background noise or un-voiced signal. This characteristic gives advantageous in voice activity detection to classify signal. In this paper we improve the performance of voice activity detection G729B using K-Nearest Neighbor method for voice or unvoiced classification. Computer simulations showed that this method has performance better than the original VAD G729B in all tested noise conditions.

3 citations


01 Jan 2010
TL;DR: Churn prediction pada telekomunikasi merupakan upaya memprediksi/mengklasifikasi pelanggan-jasa telekOMUNIKasi ying berhenti atau berpindah berlangganan dari suatu operator ke operator ying lain this article.
Abstract: Churn prediction pada pelanggan telekomunikasi merupakan upaya memprediksi/mengklasifikasi pelanggan jasa telekomunikasi yang berhenti atau berpindah berlangganan dari suatu operator ke operator yang lain. Namun dataset pada kasus churn ini biasanya memiliki kelas yang imbalance dimana jumlah instance suatu kelas (kelas active atau tidak churn atau mayor atau negatif) jauh lebih besar dari jumlah kelas yang lain (kelas churn atau minor atau positif). Akibatnya, kebanyakan classifier cenderung memprediksi kelas mayor dan mengabaikan kelas minor sehingga akurasi kelas minor sangat kecil. Salah satu pendekatan yang dilakukan untuk menangani permasalahan ini adalah dengan memodifikasi distribusi instances dari dataset yang digunakan atau yang lebih dikenal dengan pendekatan sampling-based. Teknik resampling ini meliputi oversampling, under-sampling, dan combine-sampling. Analisis yang dilakukan pada penelitian ini adalah mengetahui bagaimana pengaruh metode combine sampling yang digunakan terhadap akurasi prediksi data churn dengan melakukan penghitungan akurasi model churn prediction yang dinyatakan dalam bentuk lift curve, top decile dan gini coefficient serta f-measure untuk penghitungan akurasi prediksi data sebagai data yang imbalance. Hasil yang didapat dari penelitian menunjukkan bahwa metode combine sampling belum sesuai diterapkan pada data churn, karena cenderung masih menghasilkan nilai top decile yang kecil. Tetapi secara umum metode combine sampling ini mampu meningkatkan akurasi untuk memprediksi data minor. Dengan penerapan metode combine sampling, data churn yang memiliki tingkat imbalance yang besar dapat diklasifikasi tanpa mengorbankan data minor yang menjadi fokus penelitian. Metode combine sampling yang digunakan juga memiliki hasil evaluasi yang berbeda terhadap dataset sebagai data churn dan sebagai data imbalance.

3 citations


25 Jun 2010
TL;DR: The research concluded that DBMS give lower response time and greater throughput between those two data storage format, Native XML Database and XML.
Abstract: In the environment where information exchanges running rapidly, performance becomes a matter that is absolutely necessary to ensure the accuracy and speed of information delivery. Response time and throughput are two of the performance parameters that could be considered in choosing data storage media. A DBMS is a set of software programs that controls the organization, storage, management, and retrieval of data in a database. Along with the development of information technology, the needs of DBMS as storage media would be crucial. In addition to DBMS, there is another data storage format -had a similar structure to that relational- named XML. Nowadays Native XML Databases had been introduced as a new technology of XML. In Native XML Databases, XML files could be related one to another just like in DBMS. The objective of the research was to analyze what is best between those two data storage format. The analysis result concluded that DBMS give lower response time and greater throughput. Keywords: Native XML Database, XML, DBMS, response time, throughput

2 citations


Proceedings ArticleDOI
30 Dec 2010
TL;DR: The interoperable capabilities between the mixed WLAN standards in the environment of hybrid network are subject to be discussed and it shows that the delay bound of 802.11g DSSS-OFDM 54 Mbps is reasonable, however, it offer the highest data rate.
Abstract: The interoperable capabilities between the mixed WLAN standards in the environment of hybrid network are subject to be discussed. In this case, the extended fiber link may achieve 21.7 km. Among the modulations and the access methods applied, it shows that the delay bound of 802.11g DSSS-OFDM 54 Mbps is reasonable, however, it offer the highest data rate.

2 citations


Journal ArticleDOI
10 May 2010
TL;DR: Teknologi OMUX (Optical Multiplexer) merupakan salah satu solusi yang ditawarkan Telkom Bandung saat ini as mentioned in this paper.
Abstract: Teknologi OMUX (Optical Multiplexer) merupakan salah satu solusi yang ditawarkan Telkom Bandung saat ini. OMUX adalah suatu teknologi akses yang dikategorikan sebagai Broadband Access yang ditransmisikan melalui media fiber optik. OMUX merupakan perangkat multiplexer penjembatan layanan data dan POTS melalui media serat optik yang dibuat dengan sangat fleksibel berbentuk modul. Dalam Penelitian ini dilakukan perencanaan teknologi OMUX untuk layanan data dan POTS di Politeknik TEDC Bandung dengan cara melakukan perbandingan teknologi OMUX dengan salah satu dari teknologi pengembangan dari PON yaitu GPON (Gigabit PON). Perbandingan ini dilakukan untuk mengkaji implementasi OMUX sebagai teknologi paling tepat dan efisien diterapkan di pelanggan coorporate Telkom Bandung. Pengambilan keputusan adalah berdasarkan hasil perbandingan pengukuran parameter, analisis kebutuhan, kelebihan dan kekurangan masing-masing teknologi. Adapun parameter teknis yang dibahas adalah Power Link Budget, Rise Time Budget, spesifikasi perangkat dan Redaman. Dari hasil pengukuran parameter teknis dan analisa dapat dilihat bahwa teknologi OMUX dan GPON memenuhi persyaratan layanan broadband serta memiliki kelebihan dan kekurangan masing-masing. Nilai margin daya dan rise time jaringan menggunakan OMUX lebih besar dibandingkan dengan menggunakan GPON. Sehingga teknologi yang dipilih adalah OMUX untuk kampus Politeknik TEDC Bandung yang terletak di wilayah suburban dan tidak banyak kemungkinan pengguna akses broadband. Keputusan ini juga diambil berdasarkan pertimbangan effort yang dibutuhkan dan waktu instalasi (Time to Delivery Market).

1 citations


Proceedings ArticleDOI
TL;DR: In this paper, the authors present the application of Darboux transformation on one-dimensional Dirac equation related to the field of Quantum Information and Computation (QIC) and present the applications in cavity quantum electrodynamics and on the proposal of quantum transistor are presented.
Abstract: We present the recent works \cite{trisetyarso2011} on the application of Darboux transformation on one-dimensional Dirac equation related to the field of Quantum Information and Computation (QIC). The representation of physical system in one-dimensional equation and its transformation due to the Bagrov, Baldiotti, Gitman, and Shamshutdinova (BBGS)-Darboux transformation showing the possibility admitting the concept of relativity and the trade-off of concurrent condition of quantum and classical physics play into the area of QIC. The applications in cavity quantum electrodynamics and on the proposal of quantum transistor are presented.

01 Jan 2010
TL;DR: Sistem Komunikasi Serat Optik memungkinkan transmisi informasi dengan kapasitas data ying besar and waktu-yang cepat as mentioned in this paper.
Abstract: Sistem Komunikasi Serat Optik memungkinkan transmisi informasi dengan kapasitas data yang besar dan waktu yang cepat. Salah satu kendala pentransmisian ini dipengaruhi oleh dispersi. Berbagai metode digunakan dalam perancangan untuk mengurangi dispersi sehingga didapatkan optimalisasi transmisi, salah satunya adalah dengan perancangan Dispersion Flattened Fiber (DFF ). Dispersion Flattened Fiber merupakan serat optik yang telah dimodifikasi sehingga memungkinkan dispersi yang terjadi mendekati nol di dua atau tiga panjang gelombang yang berbeda dan sangat mendekati nol diantaranya. Perancangan DFF ini dilakukan melalui cladding serat optik yang dibuat berlapis-lapis dan dilakukan pada serat optik Single-Mode Step-Index (SM/SI). Cladding yang dibuat sampai dengan tiga lapis (Triple-Claddig).

28 Jun 2010
TL;DR: In this paper, Tweak View, Tweak Model, and Tweak Cache optima yang dilakukkan fokus pada bagaimana mengurang byte transferred.
Abstract: Seiring dengan pesatnya perkembangan aplikasi web saat ini, maka performansi aplikasi adalah isu yang sangat penting untuk diperhatikan oleh developer. Agar dapat terus memenuhi user expectation terhadap performa aplikasi tanpa menambah resource perangkat keras, maka solusi yang mungkin dilakukan adalah melakukan optimasi dari sisi scripting aplikasi. Symfony adalah sebuah framework berbasis PHP 5 yang memfasilitasi developer dengan langkah-langkah optimasi yang dinamakan Tweak Method. Tweak Method dapat dilakukan dengan 3 cara : Tweak Model, Tweak View, dan Tweak cache. Dalam studi kasus ini penulis mengimplementasikan Tweak View dan Tweak cache. Pada Tweak View, langkah optimasi yang dilakukan fokus pada bagaimana mengurangi byte transferred. Sedangkan pada Tweak cache, proses optimasi yang dilakukan adalah mengimplementasikan teknologi PHP Cache yang berfungsi untuk cache opcode dari response. Hasilnya, pengimplementasian Tweak Method pada aplikasi kasus dapat mempercepat Response Time hingga 99.26% dan byte transferred hingga 88.05% dari sebelumnya. Kata kunci : performa, optimasi, Tweak Method, Tweak View, Tweak Cache