Vanlin Sathya

Project Description: With both small-cell LTE and Wi-Fi networks available as alternatives for deployment in unlicensed bands (notably 5 GHz), the investigation into their coexistence is a topic of active interest, primarily driven by industry groups. 3GPP has recently standardized LTE-LAA that seeks to make LTE more co-existence friendly with Wi-Fi by incorporating similar sensing and back-off features. Nonetheless, the results presented by industry groups offer little consensus on important issues like respective network parameter settings that promote “fair access” as required by 3GPP. Answers to such key system deployment aspects, in turn, require credible analytical models, on which there has been a little progress to date. Accordingly, in one of the first works of its kind, we develop a new framework for estimating the throughput of Wi-Fi and LTE-LAA in coexistence scenarios via suitable modifications to the celebrated Bianchi model. The impact of various network parameters such as energy detection threshold on Wi-Fi and LTE-LAA coexistence is explored as a byproduct and corroborated via a National Instrument experimental test bed that validates the results for LTE-LAA access priority classes 1 and 3.

Project Description: With both small-cell LTE and 802.11 networks now available as alternatives for deployment in unlicensed bands at 5 GHz, investigation into their coexistence is a topic of great interest. ETSI Rel.14 has standardized LTE LAA that seeks to make LTE more coexistence friendly with Wi-Fi by incorporating listen before talk (LBT). However, the fairness of Wi-Fi and LTE-LAA sharing is a topic that has not been adequately explored. In this work, we first investigate the 3GPP definition of fair coexistence via new analytical models. By tuning the LTE-LAA parameters, we exemplify scenarios when the 3GPP notion of fairness is achieved and conversely, when not achieved. The formal notions of access and proportional fairness is then considered for these scenarios to compare and contrast with the 3GPP definition.
 

Project Description: In this work, we develop an algorithm, based only on energy detection, to detect the number of Wi-Fi basic service sets (BSSs) operating on the same channel. Such an algorithm can be used by an LTE-U base station to scale back its duty cycle when coexisting with Wi-Fi on the same channel. According to the LTE-U specification, an LTE-U BS scales back its duty cycle from 50% to 33% when it senses that the number of co-channel Wi-Fi BSSs has increased from one to two. There are two ways that this detection can be done: (a) decoding based, where the LTE-U BS decodes the Wi-Fi packet header to determine the unique Wi-Fi basic service set identification (BSSID) and (b) energy based, where only received energy levels are used to determine the number of Wi-Fi BSSs. The former approach requires an LTE-U BS to implement a Wi-Fi decoder and hence increases complexity, whereas the latter is easier to implement, requiring only an energy detector and appropriate detection thresholds to distinguish between one and two Wi-Fi APs. We analyze the latter approach and experimentally verify the feasibility of an energy detector to reliably distinguish between one and two Wi-Fi APs. In order to do so, we first experimentally determine appropriate detection thresholds using comprehensive measurements in realistic environments, both line-of-sight (LOS) and non-LOS (NLOS). These thresholds are then used to perform hypothesis based detection to devise an efficient algorithm that predicts the presence of one or two Wi-Fi BSSs. The performance of the proposed algorithm is evaluated, both theoretically and experimentally, by utilizing two metrics (a) probability of detection (PD ) and (b) probability of false alarm (PFA ). We show that using a threshold of -42 dBm delivers greater than 80% PD and less the 5% PFA which we verify both theoretically and experimentally. Hence, energy based detection is a low-complexity means of determining number of Wi-Fi BSSs to help LTE-U scale back its duty cycle appropriately.
 

Project Description: LTE-U (LTE-Unlicensed) is designed to coexist with Wi-Fi in the unlicensed band by balancing its duty cycle according to the number of coexisting Wi-Fi access points (APs) it detects. For example, a LTE-U base-station will reduce its duty cycle from 50% to 33% when it senses an increase in the number of co-channel Wi-Fi basic service sets (BSSs) from one to two. But the problem of detecting how many Wi-Fi BSS’ are operating on the channel in real-time, without decoding the Wi-Fi header, still remains. In this paper, we present a novel algorithm that solves the problem by using an auto-correlation (AC) function on the Wi-Fi preamble and setting appropriate detection thresholds to infer the number of Wi-FiBSSs operating on the channel. Performing auto-correlation on the Wi-Fi preamble is a much simpler operation than decoding the entire Wi-Fi packet, which is what would be needed ifone were to decode the MAC header to identify the BSS. We implement and experimentally validate the proposed AC detector rand demonstrate that there is a differentiable pattern of AC events between one and two Wi-Fi APs. From the collected AC events, we determine a suitable threshold for a reliable detection of Wi-Fi APs. We show that using an AC threshold of NE= 0.8, we can achieve a probability of detection (PD) of 0.9 with a probability of false alarm (PFA) of less than 0.02. Finally, we demonstrate that the performance of the proposed AC detector is superior in terms of PD and PFA compared with the energy detector (ED).
 

Project Description: The exponential increase in the number of mobile devices in use today has led to a commensurate increase in the demands on both cellular and Wi-Fi infrastructure, thus requiring that both licensed (cellular) and unlicensed (Wi-Fi) spectrum be utilized as efficiently as possible. One solution being actively pursued by industry is for cellular systems to use the unlicensed spectrum in addition to the licensed spectrum, which would require fair coexistence with Wi-Fi in the unlicensed spectrum. As per the IEEE 802.11 standard, Wi-Fi uses an energy detection (ED) threshold of -62 dBm when Long Term Evolution-Licensed Assisted Access (LTE-LAA) and/or Long Term Evolution Un-Licensed (LTE-U) nodes are deployed close by, whereas the LTE-LAA specification recommends that LTE- LAA detect Wi-Fi at -72 dBm. In our work, we evaluate the effect of this asymmetry in the ED threshold on coexistence between the two systems. We develop a coexistence simulator in ns-3 and vary both the Wi-Fi and LTE energy detection thresholds and demonstrate that lowering the Wi-Fi ED threshold from -62 dBm improves performance for both Wi-Fi and LTE-LAA. Prior work has mostly focused on determining the ED threshold that should be used by LTE-LAA/LTE-U. As far as we are aware, this is the first result that demonstrates that lowering the Wi-Fi ED threshold improves performance for both systems. The conclusion is that if Wi-Fi treats LTE-LAA/LTE-U as it would an overlapping Wi-Fi, coexistence performance improves compared to the current assumption that Wi-Fi treats LTE-LAA/LTE-U as noise.