The SF12 Well in LoRaWAN: Problem and End-Device-Based Solutions

LoRaWAN has become a popular technology for the Internet of Things (IoT) device connectivity. One of the expected properties of LoRaWAN is high network scalability. However, LoRaWAN network performance may be compromised when even a relatively small number of devices use link-layer reliability. After failed frame delivery, such devices typically tend to reduce their physical layer bit rate by increasing their spreading factor (SF). This reaction increases channel utilization, which may further degrade network performance, even into congestion collapse. When this problem arises, all the devices performing reliable frame transmission end up using SF12 (i.e., the highest SF in LoRaWAN). In this paper, we identify and characterize the described network condition, which we call the SF12 Well, in a range of scenarios and by means of extensive simulations. The results show that by using alternative SF-management techniques it is possible to avoid the problem, while achieving a packet delivery ratio increase of up to a factor of 4.7.

On Cross-Layer Interactions for Congestion Control in the Internet

Two key mechanisms of the Internet are congestion control in the Transmission Control Protocol (TCP) and Active Queue Management (AQM) in routers. The former divides the bandwidth between flows and prevents the Internet from congestion collapse. Simultaneously, the latter informs hosts of the forthcoming congestion by preventive dropping of packets in network nodes. Although these two key mechanisms may severely interact with each other, they are often being researched independently, in parallel. This has led to the development of a few new congestion controls and AQM algorithms known for excellent performance under the assumption that the counterpart remains unaltered. It is unclear, however, how these new solutions in both areas interact with each other. The purpose of this paper is to fill this gap. Namely, in an extensive set of simulations, the impact of interactions between the state-of-the-art congestion control and AQM algorithms on the TCP connection performance is studied. As a result, recommendations for using some particular TCP-AQM pairs, which are observed to perform especially well, are formulated.

Gentle Slow Start to Alleviate TCP Incast in Data Center Networks

Modern data center networks typically adopt symmetric topologies, such as leaf-spine and fat-tree. When a large number of transmission control protocol (TCP) flows in data center networks send data to the same receiver, the congestion collapse, called TCP Incast, frequently happens because of the huge packet losses and Time-Out. To address the TCP Incast issue, we firstly demonstrate that adjusting the increasing speed of the congestion window during the slow start phase is crucially important. Then we propose the Gentle Slow Start (GSS) algorithm, which adjusts the congestion window according to real-time congestion state in a gentle manner and smoothly switches from slow start to congestion avoidance phase. Furthermore, we present the implementation and design of Gentle Slow Start and also integrate it into the state-of-the-art data center transport protocols. The test results show that GSS effectively decreases the Incast probability and increases the network goodput by average 8x.

SIMULASI ANTRIAN PAKET DATA JARINGAN DENGAN MEKANISME DROP TAIL

Perkembangan penggunaan komputer dengan akses jaringan serta layanannya cepat berkembang dari masa ke masa, ini membuat kepadatan trafik data pada jaringan internet maupun intranet. Kemacetan jaringan internet pertama kali dialami pada akhir tahun 80-an, pada saat itu belum adanya mekanisme yang menangani hal tersebut. kemudian ditemukannya teorinya yaitu Congestion Avoidance and Control. Congestion adalah pengumpulan paket melebihi kapasitas bandwidth yang tersedia pada link, congestion akan mengakibatkan penurun kinerja jaringan diantaranya; multiple packet losses, utilitas link yang rendah (low throughput), delay antrian yang tinggi, dan kemacetan yang parah (congestion collapse). Penanganan kepadata jaringan sangat penting, ini membuat banyaknya metode-metode baru yang muncul dari metode sederhana sampai yang canggih, semuanya itu mempunyai kekurangan dan kelebihan, serta karakateristik masing-masing, ini menjadikan riset yang menantang untuk dipelajari dan dikembangkan, termaksud dalam penelitian ini. Pada penelitian ini dengan menggunakan simulator OPNET dibuat topologi jaringan bottlenect yang akan diimplemetasikan metode AQM klasik FIFO (Drop Tail) dengan trafik layanan seperti, FTP. Sehingga dapat dilihat penggunaan buffer pada router dalam penanganan antrian, juga berapa banyak trafik droped dan trafik sendnya, serta delay. Hasilnya dapat dilihat bahwa Drop Tail adalah solusi yang bekerja dengan baik dalam mengatasi antrian dalam buffer management dengan ditunjukkan 3 karakteristik yang baik yaitu pada Packet Dropped, Pengiriman Ulang, dan Buffer Usage.

A Dynamic Adjustment Algorithm of Slow-Start Threshold Based on RTT

The slow-start threshold in the current TCP Reno algorithm is a fixed value, and it does not change with the current network congestion status, which is prone to end the slow-start course too early or too late. Round-trip time (RTT) can reflect the current network status better. The paper presents a dynamic adjustment algorithm of slow-start threshold based on RTT, by which the value of slow-start threshold is dynamically adjusted according to the current RTT. Simulation results show that it effectively reduces congestion collapse in congestion state, and throughput is further improved under ideal network status.