Conference Papers

On the performance of social-based and location-aware forwarding strategies in urban vehicular networks

Oliveira R., Luís M., Sargento S.

Ad Hoc Networks

2019

Abstract:

High vehicular mobility in urban scenarios originates inter-vehicles communication discontinuities, a highly important factor when designing a forwarding strategy for vehicular networks. Store, carry and forward mechanisms enable the usage of vehicular networks in a large set of applications, such as sensor data collection in IoT, contributing to smart city platforms. This work evaluates the performance of several location-based and social-aware forwarding schemes through emulations and in a real scenario. Gateway Location Awareness (GLA), a location-aware ranking classification, makes use of velocity, heading angle and distance to the gateway, to select the vehicles with higher chance to deliver the information in a shorter period of time, thus differentiating nodes through their movement patterns. Aging Social-Aware Ranking (ASAR) exploits the social behavior of each vehicle, where nodes are ranked based on a historical contact table, differentiating vehicles with a high number of contacts from those who barely contact with other vehicles. To merge both location and social aforementioned algorithms, a HYBRID approach emerges, thus generating a more intelligent mechanism. For each strategy, we evaluate the influence of several parameters in the network performance, as well as we comparatively evaluate the strategies in different scenarios. Experiment results, obtained both in emulated (with real traces of both mobility and vehicular connectivity from a real city-scale urban vehicular network) and real scenarios, show the performance of GLA, ASAR and HYBRID schemes, and their results are compared to lower- and upper-bounds. The obtained results show that these strategies are a good tradeoff to maximize data delivery ratio and minimize network overhead, while making use of mobile networks as a smart city network infrastructure.

On the performance of sparse vehicular networks with road side units

Reis A.B., Sargento S., Tonguz O.K.

IEEE Vehicular Technology Conference

2011

Abstract:

The reliability of communication in vehicular networks depends mostly on the density of DSRC-enabled vehicles that form the network. In highway scenarios, and depending on the time of day, the probability of having a disconnected vehicular network can be very high, which hinders communication reliability. To improve communication in these scenarios, infrastructure points known as Road Side Units (RSU) may be used. RSUs, however, have an associated cost, and therefore the number of RSUs needs to be minimized while still providing a significant improvement on communications. In this paper we study the effect of including RSUs as relay nodes to improve communication in highway scenarios. We model the average time taken to propagate a packet to disconnected nodes (denoted as re-healing time) when considering both scenarios of connected and disconnected RSUs. We then compare the results of both these models and of a model with no RSUs. Results show significant improvements with RSU deployments, both connected and disconnected, particularly in multi-cluster communication scenarios.

On the reliability and availability of systems tolerant to stealth intrusion

Brandao L.T.A.N., Bessani A.

Proceedings - 2011 Latin-American Symposium on Dependable Computing, LADC 2011

2011

Abstract:

This paper considers the estimation of reliability and availability of intrusion-tolerant systems subject to non-detectable intrusions. Our motivation comes from the observation that typical techniques of intrusion tolerance may in certain circumstances worsen the non-functional properties they were meant to improve (e.g., dependability). We start by modeling attacks as adversarial efforts capable of affecting the intrusion rate probability of components of the system. Then, we analyze several configurations of intrusion-tolerant replication and pro-active rejuvenation, to find which ones lead to security enhancements. We analyze several parameterizations, considering different attack and rejuvenation models and taking into account the mission time of the overall system and the expected time to intrusion of its components. In doing so, we identify thresholds that distinguish between improvement and degradation. We compare the effects of replication and rejuvenation and highlight their complementarity, showing improvements of resilience not attainable with any of the techniques alone, but possible only as a synergy of their combination. We advocate the need for thorougher system models, by showing fundamental vulnerabilities arising from incomplete specifications.

On the Use of Open-Source C/C++ Static AnalysisTools in Large Projects

Pereira J.A., Vieira M.

16th European Dependable Computing Conference (EDCC 2020)

2020

Abstract:

Software applications are frequently deployed with security vulnerabilities that may open the door to attacks. In business-critical scenarios, such attacks may lead to significant financial and reputation losses. Static Analysis Tools (SATs), which analyze the source code without executing it, can be used to detect potential faults in the source code, including security vulnerabilities. However, many false alarms are normally reported, leading teams to discard the use of such tools, especially on large software projects. Existing works have dealt with the evaluation of SATs, but they are mostly based on small pieces of code designed to support the evaluation. In this paper, we present and discuss the results of the execution of two Open-Source C/C++ SATs (CPPCheck and Flawfinder) on the large open-source project Mozilla. Our goal is to study the applicability of SATs in a large project and the vulnerability categories they can detect. Results show that CppCheck could detect 83.5% of the vulnerabilities, and Flawfinder could detect 36.2%, although the number of false alarms is high (7.2% for CppCheck and 93.2% for Flawfinder). Regarding the different categories, the two SATs showed quite diverse performances (e.g., CppCheck was able to detect $92.6% of Data Protection vulnerabilities and 62.5% of Coding Practices vulnerabilities, while false alarms were 99.1% and 99.9%, respectively).

On using distributed representations of source code for the detection of C security vulnerabilities

Coimbra D., Reis S., Abreu R., Pasareanu C., Erdogmus H

arxiv

2021

Abstract:

This paper presents an evaluation of the code representation model Code2vec when trained on the task of detecting security vulnerabilities in C source code. We leverage the open-source library astminer to extract path-contexts from the abstract syntax trees of a corpus of labeled C functions. Code2vec is trained on the resulting path-contexts with the task of classifying a function as vulnerable or non-vulnerable. Using the CodeXGLUE benchmark, we show that the accuracy of Code2vec for this task is comparable to simple transformer-based methods such as pretrained RoBERTa, and outperforms more naive NLP-based methods. We achieved an accuracy of 61.43% while maintaining low computational requirements relative to larger models.

On wireless channels with M-antenna eavesdroppers: Characterization of the outage probability and outage secrecy capacity

Prabhu V.U., Rodrigues M.R.D.

GLOBECOM - IEEE Global Telecommunications Conference

2010

Abstract:

In this paper, we consider secure communications between a single antenna transmitter and a single antenna receiver in the presence of a multiple antenna eavesdropper employing MRC or SDC reception. We show that an M-antenna eavesdropper performing SDC reception has precisely the same degrading effect on secrecy as M single antenna eavesdroppers under independent fading conditions. We also show that an M-antenna eavesdropper performing MRC reception causes greater secrecy degradation than an M-antenna eavesdropper performing SDC reception or M single antenna eavesdroppers. We also derive closed-form expressions for the asymptotic high SNR outage secrecy capacity for both the MRC and the SDC eavesdropper cases. Various numerical results are presented that corroborate the analysis.

On wireless channels with m-antenna eavesdroppers: Characterization of the outage probability and ε-outage secrecy capacity

Prabhu V.U., Rodrigues M.R.D.

IEEE Transactions on Information Forensics and Security

2011

Abstract:

In this correspondence, we characterize the probability of secrecy-outage and the asymptotic high-signal-to-noise ratio (SNR) ε-outage secrecy-capacity for a single-input-single-output-multi-eavesdropper (SISOME) wireless system with eavesdroppers performing maximum ratio combining (MRC) or selection diversity combining (SDC) reception. We also consider the SISO2E case with eavesdropper antenna-correlation and finally, analyze the scenario where the eavesdropper has Rician fading links with the transmitter.

Online learning of structured predictors with multiple kernels

Martins A.F.T., Smith N.A., Xing E.P., Aguiar P.M.Q., Figueiredo M.A.T.

Journal of Machine Learning Research

2011

Abstract:

Training structured predictors often requires a considerable time selecting features or tweaking the kernel. Multiple kernel learning (MKL) sidesteps this issue by embedding the kernel learning into the training procedure. Despite the recent progress towards efficiency of MKL algorithms, the structured output case remains an open research front. We propose a family of online algorithms able to tackle variants of MKL and group-LASSO, for which we show regret, convergence, and generalization bounds. Experiments on handwriting recognition and dependency parsing attest the success of the approach.

Open-Domain Conversational Search Assistant with Transformers

Ferreira R., Leite M., Semedo D., Magalhães J.

European Conference in Information Retrieva

2021

Abstract:

Open-domain conversational search assistants aim at answering user questions about open topics in a conversational manner. In this paper we show how the Transformer architecture achieves state-of-the-art results in key IR tasks, leveraging the creation of conversational assistants that engage in open-domain conversational search with single, yet informative, answers. In particular, we propose an open-domain abstractive conversational search agent pipeline to address two major challenges: first, conversation context-aware search and second, abstractive search-answers generation. To address the first challenge, the conversation context is modeled with a query rewriting method that unfolds the context of the conversation up to a specific moment to search for the correct answers. These answers are then passed to a Transformer-based re-ranker to further improve retrieval performance. The second challenge, is tackled with recent Abstractive Transformer architectures to generate a digest of the top most relevant passages. Experiments show that Transformers deliver a solid performance across all tasks in conversational search, outperforming the best TREC CAsT 2019 baseline.