Ads supply personalization aims to balance the revenue and user engagement, two long-term objectives in social media ads, by tailoring the ad quantity and density. In the industry-scale system, the challenge for ads supply lies in modeling the counterfactual effects of a conservative supply treatment (e.g., a small density change) over an extended duration. In this paper, we present a streamlined framework for personalized ad supply. This framework optimally utilizes information from data collection policies through the doubly robust learning. Consequently, it significantly improves the accuracy of long-term treatment effect estimates. Additionally, its low-complexity design not only results in computational cost savings compared to existing methods, but also makes it scalable for billion-scale applications. Through both offline experiments and online production tests, the framework consistently demonstrated significant improvements in top-line business metrics over months. The framework has been fully deployed to live traffic in one of the world's largest social media platforms.

In many application scenarios, data consumed by real-time tasks are required to meet a maximum age, or freshness, guarantee. In this paper, we consider the end-to-end freshness constraint of data that is passed along a chain of tasks in a uniprocessor setting. We do so with few assumptions regarding the scheduling algorithm used. We present a method for selecting the periods of tasks in chains of length two and three such that the end-to-end freshness requirement is satisfied, and then extend our method to arbitrary chains. We perform evaluations of both methods using parameters from an embedded benchmark suite (E3S) and several schedulers to support our result.

The tight interaction between information technology and the physical world inherent in cyber-physical systems (CPS) can challenge traditional approaches for monitoring safety and security. Data collected for robust CPS monitoring is often sparse and may lack rich training data describing critical events/attacks. Moreover, CPS often operate in diverse environments that can have significant inter/intra-system variability. Furthermore, CPS monitors that are not robust to data sparsity and inter/intra-system variability may result in inconsistent performance and may not be trusted for monitoring safety and security. Towards overcoming these challenges, this paper presents recent work on the design of parameter-invariant (PAIN) monitors for CPS. PAIN monitors are designed such that unknown events and system variability minimally affect the monitor performance. This work describes how PAIN designs can achieve a constant false alarm rate (CFAR) in the presence of data sparsity and intra/inter system variance in real-world CPS. To demonstrate the design of PAIN monitors for safety monitoring in CPS with different types of dynamics, we consider systems with networked dynamics, linear-time invariant dynamics, and hybrid dynamics that are discussed through case studies for building actuator fault detection, meal detection in type I diabetes, and detecting hypoxia caused by pulmonary shunts in infants. In all applications, the PAIN monitor is shown to have (significantly) less variance in monitoring performance and (often) outperforms other competing approaches in the literature. Finally, an initial application of PAIN monitoring for CPS security is presented along with challenges and research directions for future security monitoring deployments.

Middleware designed for use in Distributed Real-Time and Embedded (DRE) systems enable cost and development time reductions by providing simple communications abstractions and hiding operating system-level networking API details from developers. While current middleware technologies can hide many low-level details, designers must provide a static configuration for the system's underlying network in order to achieve required performance characteristics. This has not been a problem for many types of DRE systems where the configuration of the system is relatively fixed from the factory (e.g., aircraft or naval vessels). However for truly open systems (i.e., systems where end users can add or substract components at runtime)the standard static network configuration approach cannot guarantee that required performance will be met because network resource demands are not fully known a priori. Open systems with stringent performance requirements need middleware that can dynamically manage the underlying network configuration automatically in response to changing demands. Fortunately, recent trends in networking have resulted in a wide variety of networking equipment that expose a standardized low-level interface to their configuration via the OpenFlow protocol. In this paper we discuss how OpenFlow can be leveraged by DRE middleware to automatically provide performance guarantees. In order to make the discussion concrete, we describe the architecture of our prototype middleware MIDAS as well as the details of one example network resource management strategy. We demonstrate the feasibility of our approach via performance assesment of a simple DRE application using our MIDAS and commerically available OpenFlow hardware.

We present a co-designed scheduling framework and platform architecture that together support compositional scheduling of real-time systems. The architecture is built on the Xen virtualization platform, and relies on compositional scheduling theory that uses periodic resource models as component interfaces. We implement resource models as periodic servers and consider enhancements to periodic server design that significantly improve response times of tasks and resource utilization in the system while preserving theoretical schedulability results. We present an extensive evaluation of our implementation using workloads from an avionics case study as well as synthetic ones.

In global real-time multiprocessor scheduling, a recent analysis technique for Task-level Fixed-Priority (TFP) scheduling has been shown to outperform many of the analyses for Job-level Fixed-Priority (JFP) scheduling on average. Since JFP is a generalization of TFP scheduling, and the TFP analysis technique itself has been adapted from an earlier JFP analysis, this result is counter-intuitive and in our opinion highlights the lack of good JFP scheduling techniques. Towards generalizing the superior TFP analysis to JFP scheduling, we propose the Smallest Pseudo-Deadline First (SPDF) JFP scheduling algorithm. SPDF uses a simple task-level parameter called pseudo-deadline to prioritize jobs, and hence can behave as a TFP or JFP scheduler depending on the values of the pseudodeadlines. This natural transition from TFP to JFP scheduling has enabled us to incorporate the superior TFP analysis technique in an SPDF schedulability test. We also present a pseudo-deadline assignment algorithm for SPDF scheduling that extends the well-known Optimal Priority Assignment (OPA) algorithm for TFP scheduling. We show that our algorithm is optimal for the derived schedulability test, and also present a heuristic to overcome the computational complexity issue of the optimal algorithm. Our simulation results show that the SPDF algorithm with the new analysis significantly outperforms state-of-the-art TFP and JFP analysis.

Medical cyber-physical systems (MCPS) are life-critical, context-aware, networked systems of medical devices. These systems are increasingly used in hospitals to provide high-quality continuous care for patients. The need to design complex MCPS that are both safe and effective has presented numerous challenges, including achieving high assurance in system software, intoperability, context-aware intelligence, autonomy, security and privacy, and device certifiability. In this paper, we discuss these challenges in developing MCPS, some of our work in addressing them, and several open research issues.

The hierarchical real-time scheduling framework is a widely accepted model to facilitate the design and analysis of the increasingly complex real-time systems. Interface abstraction and composition are the key issues in the hierarchical scheduling framework analysis. Schedulability is essential to guarantee that the timing requirements of all components are satisfied. In order for the design to be resource efficient, the composition must be bandwidth optimal. Associativity is desirable for open systems in which components may be added or deleted at run time. Previous techniques on compositional scheduling are either not resource efficient in some aspects, or cannot achieve optimality and associativity at the same time. In this paper, several important properties regarding the periodic resource model are identified. Based on those properties, we propose a novel interface abstraction and composition framework which achieves schedulability, optimality, and associativity. Our approach eliminates abstraction overhead in the composition.